feat: Add pipeline_uuid handling across multiple adapters and update related tests

feat: Add missing completed input lines handling in DingTalk adapter
Merge remote-tracking branch 'origin/master' into feat/card_dify_human_input
2026-07-30 08:06:07 +00:00 · 2026-07-12 19:40:59 +08:00 · 2026-07-12 18:53:13 +08:00 · 2026-07-12 18:39:48 +08:00 · 2026-07-12 00:31:12 +08:00 · 2026-07-11 22:40:08 +08:00
98 changed files with 621 additions and 4031 deletions
@@ -37,10 +37,6 @@ jobs:
        working-directory: web
        run: pnpm install --frozen-lockfile

-      - name: Run frontend unit tests
-        working-directory: web
-        run: pnpm test:unit
-
      - name: Install Playwright browsers
        working-directory: web
        run: pnpm exec playwright install --with-deps chromium
@@ -92,17 +92,6 @@ docker compose --profile all up -d

 ---

-## Live Demo
-
-**Try it now:** https://demo.langbot.dev/
-
- Email: `demo@langbot.app`
- Password: `langbot123456`
-
-_Note: Public demo environment. Do not enter sensitive information._
-
---
-
 ## Supported Platforms

 | Platform | Status | Notes |
@@ -178,6 +167,17 @@ LangBot is **agent-friendly by design** — your coding agents (Claude Code, Cod

 ---

+## Live Demo
+
+**Try it now:** https://demo.langbot.dev/
+
+- Email: `demo@langbot.app`
+- Password: `langbot123456`
+
+_Note: Public demo environment. Do not enter sensitive information._
+
+---
+
 ## Community

 [![Discord](https://img.shields.io/discord/1335141740050649118?logo=discord&label=Discord)](https://discord.gg/wdNEHETs87)
@@ -186,6 +186,12 @@ LangBot is **agent-friendly by design** — your coding agents (Claude Code, Cod

 ---

+## Star History
+
+[![Star History Chart](https://api.star-history.com/svg?repos=langbot-app/LangBot&type=Date)](https://star-history.com/#langbot-app/LangBot&Date)
+
+---
+
 ## Contributors

 Thanks to all [contributors](https://github.com/langbot-app/LangBot/graphs/contributors) who have helped make LangBot better:
@@ -92,16 +92,6 @@ docker compose --profile all up -d

 ---

-## 在线演示
-
-**立即体验：** https://demo.langbot.dev/
- 邮箱：`demo@langbot.app`
- 密码：`langbot123456`
-
-*注意：公开演示环境，请不要在其中填入任何敏感信息。*
-
---
-
 ## 支持的平台

 | 平台 | 状态 | 备注 |
@@ -180,6 +170,16 @@ docker compose --profile all up -d

 ---

+## 在线演示
+
+**立即体验：** https://demo.langbot.dev/
+- 邮箱：`demo@langbot.app`
+- 密码：`langbot123456`
+
+*注意：公开演示环境，请不要在其中填入任何敏感信息。*
+
+---
+
 ## 为 AI Agent 而生 🤖

 LangBot **从设计上就对 Agent 友好** —— 你的编码 Agent（Claude Code、Codex、Copilot、Cursor 等）可以一等公民般地操作、扩展和部署 LangBot：
@@ -203,6 +203,12 @@ LangBot **从设计上就对 Agent 友好** —— 你的编码 Agent（Claude C

 ---

+## Star 趋势
+
+[![Star History Chart](https://api.star-history.com/svg?repos=langbot-app/LangBot&type=Date)](https://star-history.com/#langbot-app/LangBot&Date)
+
+---
+
 ## 贡献者

 感谢所有[贡献者](https://github.com/langbot-app/LangBot/graphs/contributors)对 LangBot 的帮助：
@@ -91,16 +91,6 @@ docker compose --profile all up -d

 ---

-## Demo en Vivo
-
-**Pruébelo ahora:** https://demo.langbot.dev/
- Correo electrónico: `demo@langbot.app`
- Contraseña: `langbot123456`
-
-*Nota: Entorno de demostración público. No ingrese información confidencial.*
-
---
-
 ## Plataformas Soportadas

 | Plataforma | Estado | Notas |
@@ -163,6 +153,14 @@ docker compose --profile all up -d

 ---

+## Demo en Vivo
+
+**Pruébelo ahora:** https://demo.langbot.dev/
+- Correo electrónico: `demo@langbot.app`
+- Contraseña: `langbot123456`
+
+*Nota: Entorno de demostración público. No ingrese información confidencial.*
+
 ## Diseñado para Agentes de IA 🤖

 LangBot es **agent-friendly por diseño** —— tus agentes de codificación (Claude Code, Codex, Copilot, Cursor, …) pueden operar, extender y desplegar LangBot con soporte de primera clase:
@@ -184,6 +182,12 @@ LangBot es **agent-friendly por diseño** —— tus agentes de codificación (C

 ---

+## Historial de Stars
+
+[![Star History Chart](https://api.star-history.com/svg?repos=langbot-app/LangBot&type=Date)](https://star-history.com/#langbot-app/LangBot&Date)
+
+---
+
 ## Colaboradores

 Gracias a todos los [colaboradores](https://github.com/langbot-app/LangBot/graphs/contributors) que han ayudado a mejorar LangBot:
@@ -91,16 +91,6 @@ docker compose --profile all up -d

 ---

-## Démo en Ligne
-
-**Essayez maintenant :** https://demo.langbot.dev/
- Email : `demo@langbot.app`
- Mot de passe : `langbot123456`
-
-*Note : Environnement de démonstration public. Ne saisissez pas d'informations sensibles.*
-
---
-
 ## Plateformes Supportées

 | Plateforme | Statut | Notes |
@@ -163,6 +153,14 @@ docker compose --profile all up -d

 ---

+## Démo en Ligne
+
+**Essayez maintenant :** https://demo.langbot.dev/
+- Email : `demo@langbot.app`
+- Mot de passe : `langbot123456`
+
+*Note : Environnement de démonstration public. Ne saisissez pas d'informations sensibles.*
+
 ## Conçu pour les agents IA 🤖

 LangBot est **agent-friendly par conception** —— vos agents de codage (Claude Code, Codex, Copilot, Cursor, …) peuvent exploiter, étendre et déployer LangBot avec un support de premier ordre :
@@ -184,6 +182,12 @@ LangBot est **agent-friendly par conception** —— vos agents de codage (Claud

 ---

+## Historique des Stars
+
+[![Star History Chart](https://api.star-history.com/svg?repos=langbot-app/LangBot&type=Date)](https://star-history.com/#langbot-app/LangBot&Date)
+
+---
+
 ## Contributeurs

 Merci à tous les [contributeurs](https://github.com/langbot-app/LangBot/graphs/contributors) qui ont aidé à améliorer LangBot :
@@ -91,16 +91,6 @@ docker compose --profile all up -d

 ---

-## ライブデモ
-
-**今すぐ試す:** https://demo.langbot.dev/
- メール: `demo@langbot.app`
- パスワード: `langbot123456`
-
-*注意: 公開デモ環境です。機密情報を入力しないでください。*
-
---
-
 ## 対応プラットフォーム

 | プラットフォーム | ステータス | 備考 |
@@ -163,6 +153,14 @@ docker compose --profile all up -d

 ---

+## ライブデモ
+
+**今すぐ試す:** https://demo.langbot.dev/
+- メール: `demo@langbot.app`
+- パスワード: `langbot123456`
+
+*注意: 公開デモ環境です。機密情報を入力しないでください。*
+
 ## AI エージェントのために 🤖

 LangBot は **設計段階からエージェントフレンドリー** です。お使いのコーディングエージェント（Claude Code、Codex、Copilot、Cursor など）が、ファーストクラスのサポートで LangBot を操作・拡張・デプロイできます：
@@ -184,6 +182,12 @@ LangBot は **設計段階からエージェントフレンドリー** です。

 ---

+## Star 推移
+
+[![Star History Chart](https://api.star-history.com/svg?repos=langbot-app/LangBot&type=Date)](https://star-history.com/#langbot-app/LangBot&Date)
+
+---
+
 ## コントリビューター

 LangBot をより良くするために貢献してくださったすべての[コントリビューター](https://github.com/langbot-app/LangBot/graphs/contributors)に感謝します:
@@ -91,16 +91,6 @@ docker compose --profile all up -d

 ---

-## 라이브 데모
-
-**지금 체험:** https://demo.langbot.dev/
- 이메일: `demo@langbot.app`
- 비밀번호: `langbot123456`
-
-*참고: 공개 데모 환경입니다. 민감한 정보를 입력하지 마세요.*
-
---
-
 ## 지원 플랫폼

 | 플랫폼 | 상태 | 비고 |
@@ -163,6 +153,14 @@ docker compose --profile all up -d

 ---

+## 라이브 데모
+
+**지금 체험:** https://demo.langbot.dev/
+- 이메일: `demo@langbot.app`
+- 비밀번호: `langbot123456`
+
+*참고: 공개 데모 환경입니다. 민감한 정보를 입력하지 마세요.*
+
 ## AI 에이전트를 위한 설계 🤖

 LangBot은 **설계 단계부터 에이전트 친화적**입니다 —— 코딩 에이전트(Claude Code, Codex, Copilot, Cursor 등)가 일급 지원으로 LangBot을 운영·확장·배포할 수 있습니다:
@@ -184,6 +182,12 @@ LangBot은 **설계 단계부터 에이전트 친화적**입니다 —— 코딩

 ---

+## Star 추이
+
+[![Star History Chart](https://api.star-history.com/svg?repos=langbot-app/LangBot&type=Date)](https://star-history.com/#langbot-app/LangBot&Date)
+
+---
+
 ## 기여자

 LangBot을 더 나은 프로젝트로 만들어 주신 모든 [기여자](https://github.com/langbot-app/LangBot/graphs/contributors)분들께 감사드립니다:
@@ -91,16 +91,6 @@ docker compose --profile all up -d

 ---

-## Демо
-
-**Попробуйте прямо сейчас:** https://demo.langbot.dev/
- Email: `demo@langbot.app`
- Пароль: `langbot123456`
-
-*Примечание: Публичная демо-среда. Не вводите конфиденциальную информацию.*
-
---
-
 ## Поддерживаемые платформы

 | Платформа | Статус | Примечания |
@@ -163,6 +153,14 @@ docker compose --profile all up -d

 ---

+## Демо
+
+**Попробуйте прямо сейчас:** https://demo.langbot.dev/
+- Email: `demo@langbot.app`
+- Пароль: `langbot123456`
+
+*Примечание: Публичная демо-среда. Не вводите конфиденциальную информацию.*
+
 ## Создано для ИИ-агентов 🤖

 LangBot **дружелюбен к агентам по своей архитектуре** —— ваши кодинг-агенты (Claude Code, Codex, Copilot, Cursor и др.) могут управлять, расширять и развёртывать LangBot с первоклассной поддержкой:
@@ -184,6 +182,12 @@ LangBot **дружелюбен к агентам по своей архитек

 ---

+## История Stars
+
+[![Star History Chart](https://api.star-history.com/svg?repos=langbot-app/LangBot&type=Date)](https://star-history.com/#langbot-app/LangBot&Date)
+
+---
+
 ## Участники

 Спасибо всем [участникам](https://github.com/langbot-app/LangBot/graphs/contributors), которые помогли сделать LangBot лучше:
@@ -93,16 +93,6 @@ docker compose --profile all up -d

 ---

-## 線上演示
-
-**立即體驗：** https://demo.langbot.dev/
- 信箱：`demo@langbot.app`
- 密碼：`langbot123456`
-
-*注意：公開演示環境，請不要在其中填入任何敏感資訊。*
-
---
-
 ## 支援的平台

 | 平台 | 狀態 | 備註 |
@@ -179,6 +169,14 @@ docker compose --profile all up -d

 ---

+## 線上演示
+
+**立即體驗：** https://demo.langbot.dev/
+- 信箱：`demo@langbot.app`
+- 密碼：`langbot123456`
+
+*注意：公開演示環境，請不要在其中填入任何敏感資訊。*
+
 ## 為 AI Agent 而生 🤖

 LangBot **從設計上就對 Agent 友善** —— 你的編碼 Agent（Claude Code、Codex、Copilot、Cursor 等）可以一等公民般地操作、擴充和部署 LangBot：
@@ -202,6 +200,12 @@ LangBot **從設計上就對 Agent 友善** —— 你的編碼 Agent（Claude C

 ---

+## Star 趨勢
+
+[![Star History Chart](https://api.star-history.com/svg?repos=langbot-app/LangBot&type=Date)](https://star-history.com/#langbot-app/LangBot&Date)
+
+---
+
 ## 貢獻者

 感謝所有[貢獻者](https://github.com/langbot-app/LangBot/graphs/contributors)對 LangBot 的幫助：
@@ -91,16 +91,6 @@ docker compose --profile all up -d

 ---

-## Demo trực tuyến
-
-**Thử ngay:** https://demo.langbot.dev/
- Email: `demo@langbot.app`
- Mật khẩu: `langbot123456`
-
-*Lưu ý: Môi trường demo công khai. Không nhập thông tin nhạy cảm.*
-
---
-
 ## Nền tảng được hỗ trợ

 | Nền tảng | Trạng thái | Ghi chú |
@@ -163,6 +153,14 @@ docker compose --profile all up -d

 ---

+## Demo trực tuyến
+
+**Thử ngay:** https://demo.langbot.dev/
+- Email: `demo@langbot.app`
+- Mật khẩu: `langbot123456`
+
+*Lưu ý: Môi trường demo công khai. Không nhập thông tin nhạy cảm.*
+
 ## Được xây dựng cho AI Agent 🤖

 LangBot **thân thiện với agent ngay từ thiết kế** —— các coding agent của bạn (Claude Code, Codex, Copilot, Cursor, …) có thể vận hành, mở rộng và triển khai LangBot với sự hỗ trợ hạng nhất:
@@ -184,6 +182,12 @@ LangBot **thân thiện với agent ngay từ thiết kế** —— các coding

 ---

+## Lịch sử Star
+
+[![Star History Chart](https://api.star-history.com/svg?repos=langbot-app/LangBot&type=Date)](https://star-history.com/#langbot-app/LangBot&Date)
+
+---
+
 ## Người đóng góp

 Cảm ơn tất cả [người đóng góp](https://github.com/langbot-app/LangBot/graphs/contributors) đã giúp LangBot trở nên tốt hơn:
@@ -139,8 +139,7 @@ spec:
            cpu: "1000m"
        # Liveness probe to restart container if it becomes unresponsive
        livenessProbe:
-          httpGet:
-            path: /healthz
+          tcpSocket:
            port: 5400
          initialDelaySeconds: 30
          periodSeconds: 10
@@ -148,8 +147,7 @@ spec:
          failureThreshold: 3
        # Readiness probe to know when container is ready to accept traffic
        readinessProbe:
-          httpGet:
-            path: /healthz
+          tcpSocket:
            port: 5400
          initialDelaySeconds: 10
          periodSeconds: 5
@@ -267,16 +265,14 @@ spec:
            memory: "1Gi"
            cpu: "1000m"
        livenessProbe:
-          httpGet:
-            path: /healthz
+          tcpSocket:
            port: 5410
          initialDelaySeconds: 20
          periodSeconds: 10
          timeoutSeconds: 5
          failureThreshold: 3
        readinessProbe:
-          httpGet:
-            path: /healthz
+          tcpSocket:
            port: 5410
          initialDelaySeconds: 10
          periodSeconds: 5
@@ -323,10 +319,6 @@ metadata:
    app: langbot
 spec:
  replicas: 1
-  # Plugin Runtime has a single active LangBot control owner. Recreate avoids
-  # two LangBot pods fighting over that connection during a rolling update.
-  strategy:
-    type: Recreate
  selector:
    matchLabels:
      app: langbot
@@ -31,10 +31,8 @@ Both are compatible with LangBot.

 ## Installation

-Valkey Search support is included automatically on Linux and macOS. The official `valkey-glide`
-client does not currently publish a Windows package, so LangBot skips this optional dependency on
-Windows; LangBot remains usable there, but the Valkey Search backend is unavailable. To install the
-client manually on a supported platform:
+Valkey Search support is included when you install LangBot — the `valkey-glide` dependency is
+declared in `pyproject.toml`. To install manually:

 ```bash
 pip install 'valkey-glide>=2.4.1,<3.0.0'
@@ -1,6 +1,6 @@
 [project]
 name = "langbot"
-version = "4.10.6"
+version = "4.10.5"
 description = "Production-grade platform for building agentic IM bots"
 readme = "README.md"
 license-files = ["LICENSE"]
@@ -70,7 +70,7 @@ dependencies = [
    "chromadb>=1.0.0,<2.0.0",
    "qdrant-client (>=1.15.1,<2.0.0)",
    "pyseekdb==1.1.0.post3",
-    "langbot-plugin==0.4.17",
+    "langbot-plugin==0.4.13",
    "asyncpg>=0.30.0",
    "line-bot-sdk>=3.19.0",
    "matrix-nio>=0.25.2",
@@ -80,7 +80,7 @@ dependencies = [
    "pgvector>=0.4.1",
    "botocore>=1.42.39",
    "litellm>=1.0.0",
-    "valkey-glide>=2.4.1,<3.0.0; sys_platform != 'win32'", # No Windows wheels are published
+    "valkey-glide>=2.4.1,<3.0.0",
 ]
 keywords = [
    "bot",
@@ -885,10 +885,7 @@ class DingTalkClient:

        while not self._stopped:
            try:
-                # open_connection performs blocking network I/O in the DingTalk SDK.
-                # Run it off the event loop so connection stalls do not block the
-                # LangBot HTTP server and other async tasks.
-                connection = await asyncio.to_thread(self.client.open_connection)
+                connection = self.client.open_connection()

                if not connection:
                    if self.logger:
@@ -21,7 +21,7 @@ import quart

 from ... import group
 from ......utils import paths
-from ......platform.sources.websocket_manager import is_valid_session_id, ws_connection_manager
+from ......platform.sources.websocket_manager import ws_connection_manager

 logger = logging.getLogger(__name__)

@@ -203,15 +203,11 @@ class EmbedRouterGroup(group.RouterGroup):
                if session_type not in ['person', 'group']:
                    return self.http_status(400, -1, 'session_type must be person or group')

-                session_id = quart.request.args.get('session_id', '')
-                if not is_valid_session_id(session_id):
-                    return self.http_status(400, -1, 'Valid session_id is required')
-
                websocket_adapter = self.ap.platform_mgr.websocket_proxy_bot.adapter
                if not websocket_adapter:
                    return self.http_status(404, -1, 'WebSocket adapter not found')

-                messages = websocket_adapter.get_websocket_messages(pipeline_uuid, session_type, session_id)
+                messages = websocket_adapter.get_websocket_messages(pipeline_uuid, session_type)
                return self.success(data={'messages': messages})

            except Exception as e:
@@ -231,15 +227,11 @@ class EmbedRouterGroup(group.RouterGroup):
                if session_type not in ['person', 'group']:
                    return self.http_status(400, -1, 'session_type must be person or group')

-                session_id = quart.request.args.get('session_id', '')
-                if not is_valid_session_id(session_id):
-                    return self.http_status(400, -1, 'Valid session_id is required')
-
                websocket_adapter = self.ap.platform_mgr.websocket_proxy_bot.adapter
                if not websocket_adapter:
                    return self.http_status(404, -1, 'WebSocket adapter not found')

-                websocket_adapter.reset_session(pipeline_uuid, session_type, session_id)
+                websocket_adapter.reset_session(pipeline_uuid, session_type)
                return self.success(data={'message': 'Session reset successfully'})

            except Exception as e:
@@ -302,11 +294,6 @@ class EmbedRouterGroup(group.RouterGroup):
                )
                return

-            session_id = quart.websocket.args.get('session_id', '')
-            if not is_valid_session_id(session_id):
-                await quart.websocket.send(json.dumps({'type': 'error', 'message': 'Valid session_id is required'}))
-                return
-
            websocket_adapter = self.ap.platform_mgr.websocket_proxy_bot.adapter
            if not websocket_adapter:
                await quart.websocket.send(json.dumps({'type': 'error', 'message': 'WebSocket adapter not found'}))
@@ -317,7 +304,6 @@ class EmbedRouterGroup(group.RouterGroup):
                    websocket=quart.websocket._get_current_object(),
                    pipeline_uuid=pipeline_uuid,
                    session_type=session_type,
-                    session_id=session_id,
                    metadata={'user_agent': quart.websocket.headers.get('User-Agent', '')},
                )

@@ -232,13 +232,8 @@ class ModelProviderService:

        llm_models = await self.ap.llm_model_service.get_llm_models_by_provider(provider_uuid)
        embedding_models = await self.ap.embedding_models_service.get_embedding_models_by_provider(provider_uuid)
-        rerank_service = getattr(self.ap, 'rerank_models_service', None)
-        rerank_models = (
-            await rerank_service.get_rerank_models_by_provider(provider_uuid) if rerank_service is not None else []
-        )
        existing_llm_names = {model['name'] for model in llm_models}
        existing_embedding_names = {model['name'] for model in embedding_models}
-        existing_rerank_names = {model['name'] for model in rerank_models}

        filtered_models = []
        for model in scanned_models:
@@ -265,8 +260,6 @@ class ModelProviderService:
                    'already_added': (
                        model_name in existing_embedding_names
                        if scanned_type == 'embedding'
-                        else model_name in existing_rerank_names
-                        if scanned_type == 'rerank'
                        else model_name in existing_llm_names
                    ),
                }
@@ -1,7 +1,6 @@
 from __future__ import annotations

 import asyncio
-import contextlib
 import json
 import os
 import sys
@@ -29,7 +28,6 @@ _DOCKER_BOX_HOST = 'langbot_box'
 _DEFAULT_PORT = 5410

 _HEARTBEAT_INTERVAL_SEC = 20
-_HEARTBEAT_FAILURE_THRESHOLD = 3

 # Top-level keys under ``box`` that are LangBot-internal and should not be
 # forwarded to the Box runtime.
@@ -115,8 +113,6 @@ class BoxRuntimeConnector(ManagedRuntimeConnector):
        self._handler_task: asyncio.Task | None = None
        self._ctrl_task: asyncio.Task | None = None
        self._heartbeat_task: asyncio.Task | None = None
-        self._ctrl = None
-        self._generation = 0

        # Parse the relay URL once for reuse.
        parsed = urlparse(self.ws_relay_base_url)
@@ -149,67 +145,29 @@ class BoxRuntimeConnector(ManagedRuntimeConnector):
        return self.uses_websocket()

    async def initialize(self) -> None:
-        async with self._lifecycle_lock:
-            if self._closing:
-                raise BoxRuntimeUnavailableError('box runtime connector is shutting down')
-            self._generation += 1
-            await self._stop_transport()
-            try:
-                if self._uses_websocket():
-                    if platform.get_platform() == 'win32' and not self.configured_runtime_endpoint:
-                        await self._start_subprocess_then_ws()
-                    else:
-                        await self._connect_remote_ws()
-                else:
-                    await self._start_local_stdio()
-            except BaseException:
-                await self._stop_transport()
-                await self._close_managed_subprocess()
-                raise
+        if self._uses_websocket():
+            if platform.get_platform() == 'win32' and not self.configured_runtime_endpoint:
+                await self._start_subprocess_then_ws()
+            else:
+                await self._connect_remote_ws()
+        else:
+            await self._start_local_stdio()

-            if self._heartbeat_task is None or self._heartbeat_task.done():
-                self._heartbeat_task = asyncio.create_task(self._heartbeat_loop())
-
-    async def reconnect(self) -> None:
-        async with self._lifecycle_lock:
-            if self._closing:
-                raise BoxRuntimeUnavailableError('box runtime connector is shutting down')
-            self._generation += 1
-            await self._stop_transport()
-            try:
-                if self._uses_websocket():
-                    if platform.get_platform() == 'win32' and not self.configured_runtime_endpoint:
-                        await self._start_subprocess_then_ws()
-                    else:
-                        await self._connect_remote_ws()
-                else:
-                    await self._start_local_stdio()
-            except BaseException:
-                await self._stop_transport()
-                await self._close_managed_subprocess()
-                raise
-
-            if self._heartbeat_task is None or self._heartbeat_task.done():
-                self._heartbeat_task = asyncio.create_task(self._heartbeat_loop())
+        # Start heartbeat after successful connection
+        if self._heartbeat_task is None:
+            self._heartbeat_task = asyncio.create_task(self._heartbeat_loop())

    # -- heartbeat -----------------------------------------------------------

    async def _heartbeat_loop(self) -> None:
        """Periodically ping the Box runtime to detect silent disconnections."""
-        failures = 0
-        while not self._closing:
+        while True:
            await asyncio.sleep(_HEARTBEAT_INTERVAL_SEC)
            try:
                await self.ping()
-                failures = 0
                self.ap.logger.debug('Heartbeat to Box runtime success.')
            except Exception as e:
-                failures += 1
-                self.ap.logger.warning(f'Box runtime heartbeat failed ({failures}/{_HEARTBEAT_FAILURE_THRESHOLD}): {e}')
-                if failures >= _HEARTBEAT_FAILURE_THRESHOLD:
-                    failures = 0
-                    if self.runtime_disconnect_callback is not None:
-                        await self.runtime_disconnect_callback(self)
+                self.ap.logger.debug(f'Failed to heartbeat to Box runtime: {e}')

    async def ping(self) -> None:
        if self._handler is None:
@@ -238,11 +196,9 @@ class BoxRuntimeConnector(ManagedRuntimeConnector):
            # mirroring `rt -s`.
            args=['-m', 'langbot_plugin.cli.__init__', 'box', '-s', '--ws-control-port', str(self._relay_port)],
            env=env,
-            capture_stderr=False,
        )
-        self._ctrl = ctrl
        self._ctrl_task = asyncio.create_task(
-            ctrl.run(self._make_connection_callback('stdio', connected, connect_error, self._generation))
+            ctrl.run(self._make_connection_callback('stdio', connected, connect_error))
        )

        try:
@@ -326,9 +282,8 @@ class BoxRuntimeConnector(ManagedRuntimeConnector):
                await self.runtime_disconnect_callback(self)

        ctrl = WebSocketClientController(ws_url=ws_url, make_connection_failed_callback=on_connect_failed)
-        self._ctrl = ctrl
        self._ctrl_task = asyncio.create_task(
-            ctrl.run(self._make_connection_callback(transport_name, connected, connect_error, self._generation))
+            ctrl.run(self._make_connection_callback(transport_name, connected, connect_error))
        )

        try:
@@ -344,29 +299,9 @@ class BoxRuntimeConnector(ManagedRuntimeConnector):
        transport_name: str,
        connected: asyncio.Event,
        connect_error: list[Exception],
-        generation: int,
    ):
        async def new_connection_callback(connection: Connection) -> None:
-            if generation != self._generation or self._closing:
-                await connection.close()
-                return
            handler = Handler(connection)
-            connection_ready = False
-            disconnect_notified = False
-
-            async def notify_disconnect() -> None:
-                nonlocal disconnect_notified
-                if (
-                    connection_ready
-                    and not disconnect_notified
-                    and generation == self._generation
-                    and not self._closing
-                    and self.runtime_disconnect_callback is not None
-                ):
-                    disconnect_notified = True
-                    self.ap.logger.error('Disconnected from Box runtime, trying to reconnect...')
-                    await self.runtime_disconnect_callback(self)
-
            self._handler = handler
            self.client.set_handler(handler)
            self._handler_task = asyncio.create_task(handler.run())
@@ -376,74 +311,33 @@ class BoxRuntimeConnector(ManagedRuntimeConnector):
                    await handler.call_action(LangBotToBoxAction.INIT, self._filtered_box_config)
                    self.ap.logger.debug('Sent box configuration to Box runtime via INIT.')
                self.ap.logger.info(f'Connected to Box runtime via {transport_name}.')
-                connection_ready = True
                connected.set()
                await self._handler_task
-            except asyncio.CancelledError:
-                raise
            except Exception as exc:
                if not connected.is_set():
                    connect_error.append(exc)
                    connected.set()
                    return
-            finally:
-                if getattr(self, '_handler', None) is handler:
-                    self._handler = None
-                    self.client.set_handler(None)
-                await notify_disconnect()
+
+            # If we reach here, handler.run() returned normally (connection
+            # closed) or raised after the initial handshake succeeded.
+            # Either way, treat it as a disconnect.
+            if connected.is_set():
+                if self._uses_websocket():
+                    self.ap.logger.error('Disconnected from Box runtime, trying to reconnect...')
+                    if self.runtime_disconnect_callback is not None:
+                        await self.runtime_disconnect_callback(self)
+                else:
+                    self.ap.logger.error(
+                        'Disconnected from Box runtime via stdio. '
+                        'Cannot automatically reconnect — please restart LangBot.'
+                    )

        return new_connection_callback

    # -- lifecycle -----------------------------------------------------------

-    async def _stop_transport(self) -> None:
-        if self._handler is not None:
-            with contextlib.suppress(Exception):
-                await self._handler.close()
-        self.client.set_handler(None)
-        tasks = [
-            task
-            for task in (self._handler_task, self._ctrl_task)
-            if task is not None and task is not asyncio.current_task()
-        ]
-        for task in tasks:
-            task.cancel()
-        if tasks:
-            await asyncio.gather(*tasks, return_exceptions=True)
-        close_ctrl = getattr(self._ctrl, 'close', None)
-        if close_ctrl is not None:
-            with contextlib.suppress(Exception):
-                await close_ctrl()
-        self._handler = None
-        self._handler_task = None
-        self._ctrl_task = None
-        self._ctrl = None
-
-    async def aclose(self) -> None:
-        self._closing = True
-        self._generation += 1
-        if self._heartbeat_task is not None:
-            self._heartbeat_task.cancel()
-            await asyncio.gather(self._heartbeat_task, return_exceptions=True)
-            self._heartbeat_task = None
-        await self._stop_transport()
-
-        process = getattr(self, '_subprocess', None)
-        if process is not None and process.returncode is None:
-            with contextlib.suppress(ProcessLookupError):
-                process.terminate()
-            try:
-                await asyncio.wait_for(process.wait(), timeout=3)
-            except asyncio.TimeoutError:
-                with contextlib.suppress(ProcessLookupError):
-                    process.kill()
-                await process.wait()
-        self._subprocess = None
-        await self._close_managed_subprocess()
-
    def dispose(self) -> None:
-        """Best-effort synchronous compatibility wrapper; prefer ``aclose``."""
-        self._closing = True
        if self._heartbeat_task is not None:
            self._heartbeat_task.cancel()
            self._heartbeat_task = None
@@ -2,7 +2,6 @@ from __future__ import annotations

 import asyncio
 import collections
-import contextlib
 import datetime as _dt
 import enum
 import json
@@ -66,8 +65,6 @@ class BoxService:
        self.workspace_quota_mb = self._load_workspace_quota_mb()
        self._recent_errors: collections.deque[dict] = collections.deque(maxlen=_MAX_RECENT_ERRORS)
        self._shutdown_task = None
-        self._reconnect_task: asyncio.Task | None = None
-        self._closing = False
        self._available = False
        self._connector_error: str = ''
        self._reconnecting = False
@@ -113,8 +110,6 @@ class BoxService:
            self.ap.logger.warning(f'LangBot Box runtime unavailable, sandbox features disabled: {exc}')
            self._available = False
            self._connector_error = str(exc)
-            if self._runtime_connector is not None:
-                await self._on_runtime_disconnect(self._runtime_connector)

    async def _on_runtime_disconnect(self, connector: BoxRuntimeConnector) -> None:
        """Called by the connector when the Box runtime connection drops.
@@ -123,34 +118,29 @@ class BoxService:
        Skipped entirely when Box is disabled by config — that path should
        never have connected in the first place.
        """
-        if not self._enabled or self._closing:
+        if not self._enabled:
            return
-        if self._reconnect_task is not None and not self._reconnect_task.done():
+        if self._reconnecting:
            return  # Another reconnect loop is already running
        self._reconnecting = True
        self._available = False
        self._connector_error = 'Disconnected from Box runtime'
        self.ap.logger.warning('Box runtime disconnected, sandbox features temporarily disabled.')
-        self._reconnect_task = asyncio.create_task(self._reconnect_loop(connector))
+        asyncio.create_task(self._reconnect_loop(connector))

    async def _reconnect_loop(self, connector: BoxRuntimeConnector) -> None:
        """Retry reconnection with exponential backoff (3s → 60s max)."""
        delay = 3
        max_delay = 60
        try:
-            while not self._closing:
+            while True:
                self.ap.logger.info(f'Attempting to reconnect to Box runtime in {delay}s...')
                await asyncio.sleep(delay)
                try:
-                    await connector.reconnect()
-                    self._ensure_default_workspace()
-                    await self._purge_attachment_dirs()
+                    connector.dispose()
+                    await connector.initialize()
                    self._available = True
                    self._connector_error = ''
-                    skill_mgr = getattr(self.ap, 'skill_mgr', None)
-                    reload_skills = getattr(skill_mgr, 'reload_skills', None)
-                    if callable(reload_skills):
-                        await reload_skills()
                    self.ap.logger.info('Box runtime reconnected, sandbox features restored.')
                    return
                except Exception as exc:
@@ -159,7 +149,6 @@ class BoxService:
                    delay = min(delay * 2, max_delay)
        finally:
            self._reconnecting = False
-            self._reconnect_task = None

    @property
    def available(self) -> bool:
@@ -849,28 +838,14 @@ class BoxService:
        return attachments

    async def shutdown(self):
-        if self._closing:
-            return
-        self._closing = True
-        self._available = False
-        reconnect_task = self._reconnect_task
-        self._reconnect_task = None
-        if reconnect_task is not None and reconnect_task is not asyncio.current_task():
-            reconnect_task.cancel()
-            await asyncio.gather(reconnect_task, return_exceptions=True)
-        # The runtime may already be offline. A failed best-effort SHUTDOWN RPC
-        # must not prevent us from cancelling transports and reaping children.
-        with contextlib.suppress(Exception):
-            await self.client.shutdown()
-        if self._runtime_connector is not None:
-            await self._runtime_connector.aclose()
+        await self.client.shutdown()

    def dispose(self):
+        if self._runtime_connector is not None:
+            self._runtime_connector.dispose()
        loop = getattr(self.ap, 'event_loop', None)
        if loop is not None and not loop.is_closed() and (self._shutdown_task is None or self._shutdown_task.done()):
            self._shutdown_task = loop.create_task(self.shutdown())
-        elif self._runtime_connector is not None:
-            self._runtime_connector.dispose()

    async def get_sessions(self) -> list[dict]:
        if not self._available:
@@ -4,7 +4,6 @@ import logging
 import asyncio
 import traceback
 import os
-import contextlib

 from ..platform import botmgr as im_mgr
 from ..platform.webhook_pusher import WebhookPusher
@@ -167,8 +166,7 @@ class Application:
    maintenance_service: maintenance_service.MaintenanceService = None

    def __init__(self):
-        self._shutdown_lock = asyncio.Lock()
-        self._shutdown_complete = False
+        pass

    async def initialize(self):
        pass
@@ -320,39 +318,7 @@ class Application:
            return default
        return parsed

-    async def shutdown(self):
-        """Stop application work and deterministically release runtime resources."""
-        async with self._shutdown_lock:
-            if self._shutdown_complete:
-                return
-
-            if self.task_mgr is not None:
-                self.task_mgr.cancel_by_scope(core_entities.LifecycleControlScope.APPLICATION)
-            if self.platform_mgr is not None:
-                with contextlib.suppress(Exception):
-                    await self.platform_mgr.shutdown()
-            if self.tool_mgr is not None:
-                with contextlib.suppress(Exception):
-                    await self.tool_mgr.shutdown()
-            if self.box_service is not None:
-                with contextlib.suppress(Exception):
-                    await self.box_service.shutdown()
-            if self.plugin_connector is not None:
-                with contextlib.suppress(Exception):
-                    await self.plugin_connector.aclose()
-
-            if self.task_mgr is not None:
-                tasks = [wrapper.task for wrapper in self.task_mgr.tasks if not wrapper.task.done()]
-                if tasks:
-                    await asyncio.gather(*tasks, return_exceptions=True)
-            self._shutdown_complete = True
-
    def dispose(self):
-        """Compatibility wrapper for callers that cannot await shutdown."""
-        loop = self.event_loop
-        if loop is not None and not loop.is_closed():
-            loop.create_task(self.shutdown())
-            return
        if self.plugin_connector is not None:
            self.plugin_connector.dispose()
        if self.box_service is not None:
@@ -46,38 +46,21 @@ async def make_app(loop: asyncio.AbstractEventLoop) -> app.Application:

 async def main(loop: asyncio.AbstractEventLoop):
    app_inst: app.Application | None = None
-    runtime_loop = asyncio.get_running_loop()
-    shutdown_requested = asyncio.Event()
-    run_task: asyncio.Task | None = None
    try:
+        # Hang system signal processing
        import signal

        def signal_handler(sig, frame):
+            if app_inst is not None:
+                app_inst.dispose()
            print('[Signal] Program exit.')
-            runtime_loop.call_soon_threadsafe(shutdown_requested.set)
+            os._exit(0)

        signal.signal(signal.SIGINT, signal_handler)
-        if hasattr(signal, 'SIGTERM'):
-            signal.signal(signal.SIGTERM, signal_handler)

        app_inst = await make_app(loop)
-        if app_inst is None:
-            return
-        run_task = asyncio.create_task(app_inst.run())
-        shutdown_task = asyncio.create_task(shutdown_requested.wait())
-        done, pending = await asyncio.wait((run_task, shutdown_task), return_when=asyncio.FIRST_COMPLETED)
-        if shutdown_task in done:
-            await app_inst.shutdown()
-            if not run_task.done():
-                run_task.cancel()
-        for task in pending:
-            task.cancel()
-        results = await asyncio.gather(run_task, shutdown_task, return_exceptions=True)
-        run_result = results[0]
-        if isinstance(run_result, BaseException) and not isinstance(run_result, asyncio.CancelledError):
-            raise run_result
+        await app_inst.run()
    except Exception:
-        traceback.print_exc()
-    finally:
        if app_inst is not None:
-            await app_inst.shutdown()
+            app_inst.dispose()
+        traceback.print_exc()
@@ -1,5 +1,7 @@
 from __future__ import annotations

+import asyncio
+
 from .. import stage, app
 from ...utils import version, proxy
 from ...pipeline import pool, controller, pipelinemgr
@@ -185,16 +187,11 @@ class BuildAppStage(stage.BootingStage):
        ap.maintenance_service = maintenance_service_inst

        async def runtime_disconnect_callback(connector: plugin_connector.PluginRuntimeConnector) -> None:
-            connector.schedule_reconnect()
+            await asyncio.sleep(3)
+            await plugin_connector_inst.initialize()

        plugin_connector_inst = plugin_connector.PluginRuntimeConnector(ap, runtime_disconnect_callback)
-        try:
-            await plugin_connector_inst.initialize()
-        except Exception as exc:
-            # Keep the API/UI available while an external or managed runtime is
-            # starting, then recover in the background with bounded backoff.
-            ap.logger.warning(f'Plugin runtime unavailable during startup; reconnecting in background: {exc}')
-            plugin_connector_inst.schedule_reconnect()
+        await plugin_connector_inst.initialize()
        ap.plugin_connector = plugin_connector_inst

        ctrl = controller.Controller(ap)
@@ -1,12 +1,10 @@
 from __future__ import annotations
 import typing

-import sqlalchemy

 from .. import handler
 from ... import entities
 from ... import plugin_diagnostics
-from ....entity.persistence.bot import BotAdmin
 import langbot_plugin.api.entities.builtin.provider.message as provider_message
 import langbot_plugin.api.entities.builtin.provider.session as provider_session
 import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
@@ -26,14 +24,7 @@ class CommandHandler(handler.MessageHandler):

        privilege = 1

-        admins = await self.ap.persistence_mgr.execute_async(
-            sqlalchemy.select(BotAdmin).where(
-                BotAdmin.bot_uuid == (query.bot_uuid or ''),
-                BotAdmin.launcher_type == query.launcher_type.value,
-                BotAdmin.launcher_id == str(query.launcher_id),
-            )
-        )
-        if admins.first() is not None:
+        if f'{query.launcher_type.value}_{query.launcher_id}' in self.ap.instance_config.data['admins']:
            privilege = 2

        spt = command_text.split(' ')
@@ -240,11 +240,7 @@ class AiocqhttpMessageConverter(abstract_platform_adapter.AbstractMessageConvert
        async def process_message_data(msg_data, reply_list):
            if msg_data['type'] == 'image':
                image_base64, image_format = await image.qq_image_url_to_base64(msg_data['data']['url'])
-                reply_list.append(
-                    platform_message.Image(
-                        url=msg_data['data']['url'], base64=f'data:image/{image_format};base64,{image_base64}'
-                    )
-                )
+                reply_list.append(platform_message.Image(base64=f'data:image/{image_format};base64,{image_base64}'))

            elif msg_data['type'] == 'text':
                reply_list.append(platform_message.Plain(text=msg_data['data']['text']))
@@ -289,9 +285,7 @@ class AiocqhttpMessageConverter(abstract_platform_adapter.AbstractMessageConvert
                    image_msg = platform_message.Face(face_id=face_id, face_name=face_name)
                else:
                    image_base64, image_format = await image.qq_image_url_to_base64(msg.data['url'])
-                    image_msg = platform_message.Image(
-                        url=msg.data['url'], base64=f'data:image/{image_format};base64,{image_base64}'
-                    )
+                    image_msg = platform_message.Image(base64=f'data:image/{image_format};base64,{image_base64}')
                yiri_msg_list.append(image_msg)
            elif msg.type == 'forward':
                # 暂时不太合理
@@ -497,11 +491,7 @@ class AiocqhttpAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter)
    message_converter: AiocqhttpMessageConverter = AiocqhttpMessageConverter()
    event_converter: AiocqhttpEventConverter = pydantic.Field(default_factory=AiocqhttpEventConverter)

-    on_websocket_connection_event_cache: list[aiocqhttp.Event] = []
-    _listener_wrappers: dict[
-        tuple[typing.Type[platform_events.Event], typing.Callable],
-        tuple[str, typing.Callable],
-    ] = {}
+    on_websocket_connection_event_cache: typing.List[typing.Callable[[aiocqhttp.Event], None]] = []

    def __init__(self, config: dict, logger: abstract_platform_logger.AbstractEventLogger):
        super().__init__(
@@ -516,7 +506,6 @@ class AiocqhttpAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter)
        self.config['shutdown_trigger'] = shutdown_trigger_placeholder

        self.on_websocket_connection_event_cache = []
-        self._listener_wrappers = {}

        if 'access-token' in config:
            self.bot = aiocqhttp.CQHttp(access_token=config['access-token'])
@@ -524,16 +513,6 @@ class AiocqhttpAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter)
        else:
            self.bot = aiocqhttp.CQHttp()

-        self.bot.on_websocket_connection(self._on_websocket_connection)
-
-    async def _on_websocket_connection(self, event: aiocqhttp.Event):
-        for cached_event in self.on_websocket_connection_event_cache:
-            if cached_event.self_id == event.self_id and cached_event.time == event.time:
-                return
-
-        self.on_websocket_connection_event_cache.append(event)
-        await self.logger.info(f'WebSocket connection established, bot id: {event.self_id}')
-
    async def send_message(self, target_type: str, target_id: str, message: platform_message.MessageChain):
        # Check if message contains a Forward component
        forward_msg = message.get_first(platform_message.Forward)
@@ -669,14 +648,22 @@ class AiocqhttpAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter)

        if event_type == platform_events.GroupMessage:
            self.bot.on_message('group')(on_message)
-            self._listener_wrappers[(event_type, callback)] = ('message.group', on_message)
            # self.bot.on_notice()(on_message)
        elif event_type == platform_events.FriendMessage:
            self.bot.on_message('private')(on_message)
-            self._listener_wrappers[(event_type, callback)] = ('message.private', on_message)
            # self.bot.on_notice()(on_message)
        # print(event_type)

+        async def on_websocket_connection(event: aiocqhttp.Event):
+            for event in self.on_websocket_connection_event_cache:
+                if event.self_id == event.self_id and event.time == event.time:
+                    return
+
+            self.on_websocket_connection_event_cache.append(event)
+            await self.logger.info(f'WebSocket connection established, bot id: {event.self_id}')
+
+        self.bot.on_websocket_connection(on_websocket_connection)
+
    def unregister_listener(
        self,
        event_type: typing.Type[platform_events.Event],
@@ -684,12 +671,7 @@ class AiocqhttpAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter)
            [platform_events.Event, abstract_platform_adapter.AbstractMessagePlatformAdapter], None
        ],
    ):
-        listener = self._listener_wrappers.pop((event_type, callback), None)
-        if listener is None:
-            return
-
-        event_name, wrapper = listener
-        self.bot._bus.unsubscribe(event_name, wrapper)
+        return super().unregister_listener(event_type, callback)

    async def run_async(self):
        await self.bot._server_app.run_task(**self.config)
@@ -783,9 +783,7 @@ class DiscordMessageConverter(abstract_platform_adapter.AbstractMessageConverter
                image_data = await response.read()
                image_base64 = base64.b64encode(image_data).decode('utf-8')
                image_format = response.headers['Content-Type']
-                element_list.append(
-                    platform_message.Image(url=attachment.url, base64=f'data:{image_format};base64,{image_base64}')
-                )
+                element_list.append(platform_message.Image(base64=f'data:{image_format};base64,{image_base64}'))

        return platform_message.MessageChain(element_list)

@@ -242,33 +242,6 @@ def _lark_extract_action_form_inputs(action: typing.Any, action_value_obj: dict)
    return form_inputs


-class NonBlockingLarkWSClient(lark_oapi.ws.Client):
-    """Keep the SDK's synchronous connection lookup off LangBot's event loop.
-
-    lark-oapi performs ``requests.post`` inside its async ``_connect`` method.
-    A stalled TLS handshake therefore freezes Quart and every other adapter in
-    the process. Pre-fetch the URL in a worker thread, then let the SDK finish
-    the WebSocket setup on the main loop with the already-resolved URL.
-    """
-
-    def __init__(self, *args, **kwargs):
-        super().__init__(*args, **kwargs)
-        self._langbot_connect_lock = asyncio.Lock()
-
-    async def _connect(self) -> None:
-        async with self._langbot_connect_lock:
-            if self._conn is not None:
-                return
-
-            conn_url = await asyncio.to_thread(self._get_conn_url)
-            original_get_conn_url = self._get_conn_url
-            self._get_conn_url = lambda: conn_url
-            try:
-                await super()._connect()
-            finally:
-                self._get_conn_url = original_get_conn_url
-
-
 class AESCipher(object):
    def __init__(self, key):
        self.bs = AES.block_size
@@ -1267,9 +1240,7 @@ class LarkAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
        bot_account_id = config['bot_name']

        domain = self._resolve_domain(config)
-        bot = NonBlockingLarkWSClient(
-            config['app_id'], config['app_secret'], event_handler=event_handler, domain=domain
-        )
+        bot = lark_oapi.ws.Client(config['app_id'], config['app_secret'], event_handler=event_handler, domain=domain)
        api_client = self.build_api_client(config)
        cipher = AESCipher(config.get('encrypt-key', ''))
        self.request_app_ticket(api_client, config)
@@ -1436,27 +1407,6 @@ class LarkAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):
            )
        return api_client

-    @staticmethod
-    def _has_markdown_table(text_elements: list) -> bool:
-        """Check if text elements contain markdown table syntax (|...|).
-
-        A markdown table requires:
-        - At least one line with pipe characters (|)
-        - A separator line with pipes and dashes (|---|)
-        """
-        # Regex to detect markdown table: lines with pipes and separator lines with dashes
-        table_pattern = re.compile(r'^\s*\|.+\|\s*$', re.MULTILINE)
-        separator_pattern = re.compile(r'^\s*\|\s*-+(\s*\|\s*-+)*\s*\|\s*$', re.MULTILINE)
-
-        for paragraph in text_elements:
-            for ele in paragraph:
-                if ele.get('tag') == 'md':
-                    text = ele.get('text', '')
-                    # Quick check: if has pipes and separator line, it's likely a table
-                    if '|' in text and table_pattern.search(text) and separator_pattern.search(text):
-                        return True
-        return False
-
    async def send_message(self, target_type: str, target_id: str, message: platform_message.MessageChain):
        text_elements, media_items = await self.message_converter.yiri2target(message, self.api_client)

@@ -1470,10 +1420,7 @@ class LarkAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):

        # Send text message if there are text elements
        if text_elements:
-            # Use 'post' format if: has @mentions OR has markdown tables
-            has_at = any(ele['tag'] == 'at' for paragraph in text_elements for ele in paragraph)
-            has_table = self._has_markdown_table(text_elements)
-            needs_post = has_at or has_table
+            needs_post = any(ele['tag'] == 'at' for paragraph in text_elements for ele in paragraph)

            if needs_post:
                msg_type = 'post'
@@ -1967,10 +1914,9 @@ class LarkAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter):

        # Send text message if there are text elements
        if text_elements:
-            # Use 'post' format if: has @mentions OR has markdown tables
-            has_at = any(ele['tag'] == 'at' for paragraph in text_elements for ele in paragraph)
-            has_table = self._has_markdown_table(text_elements)
-            needs_post = has_at or has_table
+            # Determine msg_type based on content: use 'post' if at mentions
+            # are present (requires post paragraph structure), otherwise 'text'
+            needs_post = any(ele['tag'] == 'at' for paragraph in text_elements for ele in paragraph)

            if needs_post:
                msg_type = 'post'
@@ -102,7 +102,7 @@ class QQOfficialMessageConverter(abstract_platform_adapter.AbstractMessageConver
        yiri_msg_list.append(platform_message.Source(id=message_id, time=datetime.datetime.now()))
        if pic_url is not None:
            base64_url = await image.get_qq_official_image_base64(pic_url=pic_url, content_type=content_type)
-            yiri_msg_list.append(platform_message.Image(url=pic_url, base64=base64_url))
+            yiri_msg_list.append(platform_message.Image(base64=base64_url))

        yiri_msg_list.append(platform_message.Plain(text=message))
        chain = platform_message.MessageChain(yiri_msg_list)
@@ -47,7 +47,7 @@ class SlackMessageConverter(abstract_platform_adapter.AbstractMessageConverter):
        yiri_msg_list.append(platform_message.Source(id=message_id, time=datetime.datetime.now()))
        if pic_url is not None:
            base64_url = await image.get_slack_image_to_base64(pic_url=pic_url, bot_token=bot.bot_token)
-            yiri_msg_list.append(platform_message.Image(url=pic_url, base64=base64_url))
+            yiri_msg_list.append(platform_message.Image(base64=base64_url))

        yiri_msg_list.append(platform_message.Plain(text=message))
        chain = platform_message.MessageChain(yiri_msg_list)
@@ -157,8 +157,7 @@ class TelegramMessageConverter(abstract_platform_adapter.AbstractMessageConverte

            message_components.append(
                platform_message.Image(
-                    url=file.file_path,
-                    base64=f'data:{file_format};base64,{base64.b64encode(file_bytes).decode("utf-8")}',
+                    base64=f'data:{file_format};base64,{base64.b64encode(file_bytes).decode("utf-8")}'
                )
            )

@@ -13,7 +13,7 @@ import langbot_plugin.api.entities.builtin.platform.events as platform_events
 import langbot_plugin.api.entities.builtin.platform.entities as platform_entities
 import langbot_plugin.api.definition.abstract.platform.event_logger as abstract_platform_logger
 from ...core import app
-from .websocket_manager import WebSocketConnection, is_valid_session_id, ws_connection_manager
+from .websocket_manager import ws_connection_manager, WebSocketConnection

 logger = logging.getLogger(__name__)

@@ -91,59 +91,6 @@ class WebSocketAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter)
        self.outbound_message_queue = asyncio.Queue()
        self.stream_enabled = True

-    @staticmethod
-    def _conversation_key(pipeline_uuid: str, session_id: str | None = None) -> str:
-        """Return the history key for a pipeline/client conversation."""
-        return f'{pipeline_uuid}:{session_id}' if session_id else pipeline_uuid
-
-    @staticmethod
-    def _parse_embed_target(target_id: str) -> tuple[str, str] | None:
-        """Extract pipeline and session identifiers from a stable embed launcher."""
-        target_value = str(target_id)
-        for prefix in ('websocket_', 'websocketgroup_'):
-            if target_value.startswith(prefix):
-                target = target_value[len(prefix) :]
-                break
-        else:
-            return None
-        if ':' not in target:
-            return None
-        pipeline_uuid, session_id = target.rsplit(':', 1)
-        if not pipeline_uuid or not is_valid_session_id(session_id):
-            return None
-        return pipeline_uuid, session_id
-
-    @classmethod
-    async def _get_connection_from_target(cls, target_id: str):
-        """Resolve a person or group WebSocket launcher to its connection."""
-        target_value = str(target_id)
-        for prefix in ('websocket_', 'websocketgroup_'):
-            if target_value.startswith(prefix):
-                target = target_value[len(prefix) :]
-                break
-        else:
-            return None
-        connection = await ws_connection_manager.get_connection(target)
-        if connection is not None:
-            return connection
-        embed_target = cls._parse_embed_target(target_id)
-        if embed_target is not None:
-            pipeline_uuid, session_id = embed_target
-            return await ws_connection_manager.get_connection_by_session_id(session_id, pipeline_uuid)
-        return await ws_connection_manager.get_connection_by_session_id(target)
-
-    async def _get_message_context(self, message_source) -> tuple[str, str | None]:
-        """Resolve the originating pipeline and browser session for a reply."""
-        sender = getattr(message_source, 'sender', None)
-        sender_id = getattr(sender, 'id', '')
-        connection = await self._get_connection_from_target(sender_id)
-        if connection is not None:
-            return connection.pipeline_uuid, connection.session_id
-        embed_target = self._parse_embed_target(sender_id)
-        if embed_target is not None:
-            return embed_target
-        return typing.cast(str, self.ap.platform_mgr.websocket_proxy_bot.bot_entity.use_pipeline_uuid), None
-
    async def send_message(
        self,
        target_type: str,
@@ -156,26 +103,15 @@ class WebSocketAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter)
        target_id 可能是 launcher_id（如 websocket_xxx）或 pipeline_uuid。
        我们需要尝试两种方式来确保消息能够送达。
        """
-        connection = await self._get_connection_from_target(target_id)
-        if connection is not None:
-            pipeline_uuid = connection.pipeline_uuid
-            session_id = connection.session_id
-        else:
-            embed_target = self._parse_embed_target(target_id)
-            if embed_target is not None:
-                pipeline_uuid, session_id = embed_target
-            else:
-                pipeline_uuid = typing.cast(
-                    str,
-                    self.ap.platform_mgr.websocket_proxy_bot.bot_entity.use_pipeline_uuid,
-                )
-                session_id = None
+        # 获取当前的 pipeline_uuid
+        pipeline_uuid = self.ap.platform_mgr.websocket_proxy_bot.bot_entity.use_pipeline_uuid
        session_type = 'group' if target_type == 'group' else 'person'
-        conversation_key = self._conversation_key(pipeline_uuid, session_id)

+        # 选择会话
        session = self.websocket_group_session if session_type == 'group' else self.websocket_person_session

-        msg_id = len(session.get_message_list(conversation_key)) + 1
+        # 生成唯一消息ID
+        msg_id = len(session.get_message_list(pipeline_uuid)) + 1

        message_data = WebSocketMessage(
            id=msg_id,
@@ -186,8 +122,10 @@ class WebSocketAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter)
            is_final=True,
        )

-        session.get_message_list(conversation_key).append(message_data)
+        # 保存到历史记录
+        session.get_message_list(pipeline_uuid).append(message_data)

+        # 直接广播到当前pipeline的连接
        await ws_connection_manager.broadcast_to_pipeline(
            pipeline_uuid,
            {
@@ -196,7 +134,6 @@ class WebSocketAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter)
                'data': message_data.model_dump(),
            },
            session_type=session_type,
-            session_id=session_id,
        )

        return message_data.model_dump()
@@ -215,11 +152,12 @@ class WebSocketAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter)
            else self.websocket_person_session
        )

-        pipeline_uuid, session_id = await self._get_message_context(message_source)
+        # 从message_source获取pipeline_uuid和connection_id
+        pipeline_uuid = self.ap.platform_mgr.websocket_proxy_bot.bot_entity.use_pipeline_uuid
        session_type = 'group' if isinstance(message_source, platform_events.GroupMessage) else 'person'
-        conversation_key = self._conversation_key(pipeline_uuid, session_id)

-        msg_id = len(session.get_message_list(conversation_key)) + 1
+        # 生成新的消息ID
+        msg_id = len(session.get_message_list(pipeline_uuid)) + 1

        message_data = WebSocketMessage(
            id=msg_id,
@@ -230,8 +168,10 @@ class WebSocketAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter)
            is_final=True,
        )

-        session.get_message_list(conversation_key).append(message_data)
+        # 保存到历史记录
+        session.get_message_list(pipeline_uuid).append(message_data)

+        # 直接广播到所有该pipeline的连接，包含session_type信息
        await ws_connection_manager.broadcast_to_pipeline(
            pipeline_uuid,
            {
@@ -240,7 +180,6 @@ class WebSocketAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter)
                'data': message_data.model_dump(),
            },
            session_type=session_type,
-            session_id=session_id,
        )

        return message_data.model_dump()
@@ -261,11 +200,10 @@ class WebSocketAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter)
            else self.websocket_person_session
        )

-        pipeline_uuid, session_id = await self._get_message_context(message_source)
+        pipeline_uuid = self.ap.platform_mgr.websocket_proxy_bot.bot_entity.use_pipeline_uuid
        session_type = 'group' if isinstance(message_source, platform_events.GroupMessage) else 'person'
-        conversation_key = self._conversation_key(pipeline_uuid, session_id)
-        message_list = session.get_message_list(conversation_key)
-        stream_message_indexes = session.get_stream_message_indexes(conversation_key)
+        message_list = session.get_message_list(pipeline_uuid)
+        stream_message_indexes = session.get_stream_message_indexes(pipeline_uuid)

        # Streaming messages in LangBot have a stable resp_message_id during the same assistant reply.
        # Use it as the primary key to avoid overwriting an old card from a previous reply.
@@ -309,6 +247,7 @@ class WebSocketAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter)
        if message_is_final and resp_message_id:
            stream_message_indexes.pop(resp_message_id, None)

+        # 直接广播到所有该pipeline的连接，包含session_type信息
        await ws_connection_manager.broadcast_to_pipeline(
            pipeline_uuid,
            {
@@ -317,7 +256,6 @@ class WebSocketAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter)
                'data': message_data.model_dump(),
            },
            session_type=session_type,
-            session_id=session_id,
        )

        return message_data.model_dump()
@@ -443,19 +381,23 @@ class WebSocketAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter)
        """
        pipeline_uuid = connection.pipeline_uuid
        session_type = connection.session_type
-        conversation_key = self._conversation_key(pipeline_uuid, connection.session_id)

+        # 获取stream参数，默认为True
        self.stream_enabled = message_data.get('stream', True)

+        # 选择会话
        use_session = self.websocket_group_session if session_type == 'group' else self.websocket_person_session

+        # 解析消息链
        message_chain_obj = message_data.get('message', [])

+        # 处理图片组件：将path转换为base64
        await self._process_image_components(message_chain_obj)

        message_chain = platform_message.MessageChain.model_validate(message_chain_obj)

-        message_id = len(use_session.get_message_list(conversation_key)) + 1
+        # 生成消息ID
+        message_id = len(use_session.get_message_list(pipeline_uuid)) + 1

        # 保存用户消息
        user_message = WebSocketMessage(
@@ -467,8 +409,9 @@ class WebSocketAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter)
            connection_id=connection.connection_id,
            is_final=True,  # 用户消息始终是完整的，非流式
        )
-        use_session.get_message_list(conversation_key).append(user_message)
+        use_session.get_message_list(pipeline_uuid).append(user_message)

+        # 广播用户消息到所有连接（包括发送者），包含session_type信息
        await ws_connection_manager.broadcast_to_pipeline(
            pipeline_uuid,
            {
@@ -477,27 +420,25 @@ class WebSocketAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter)
                'data': user_message.model_dump(),
            },
            session_type=session_type,
-            session_id=connection.session_id,
        )

        # 添加消息源
        message_chain.insert(0, platform_message.Source(id=message_id, time=datetime.now().timestamp()))

        # 创建事件
-        launcher_id = f'{pipeline_uuid}:{connection.session_id}' if connection.session_id else connection.connection_id
        if session_type == 'person':
-            sender = platform_entities.Friend(id=f'websocket_{launcher_id}', nickname='User', remark='User')
+            sender = platform_entities.Friend(
+                id=f'websocket_{connection.connection_id}', nickname='User', remark='User'
+            )
            event = platform_events.FriendMessage(
                sender=sender, message_chain=message_chain, time=datetime.now().timestamp()
            )
        else:
            group = platform_entities.Group(
-                id=f'websocketgroup_{launcher_id}' if connection.session_id else 'websocketgroup',
-                name='Group',
-                permission=platform_entities.Permission.Member,
+                id='websocketgroup', name='Group', permission=platform_entities.Permission.Member
            )
            sender = platform_entities.GroupMember(
-                id=f'websocket_{launcher_id}',
+                id=f'websocket_{connection.connection_id}',
                member_name='User',
                group=group,
                permission=platform_entities.Permission.Member,
@@ -527,47 +468,22 @@ class WebSocketAdapter(abstract_platform_adapter.AbstractMessagePlatformAdapter)
        if event.__class__ in listeners:
            asyncio.create_task(listeners[event.__class__](event, callback_adapter))

-    def get_websocket_messages(
-        self,
-        pipeline_uuid: str,
-        session_type: str,
-        session_id: str | None = None,
-    ) -> list[dict]:
-        """Return history for one pipeline/client conversation."""
-        conversation_key = self._conversation_key(pipeline_uuid, session_id)
-        session = self.websocket_person_session if session_type == 'person' else self.websocket_group_session
-        return [message.model_dump() for message in session.message_lists.get(conversation_key, [])]
+    def get_websocket_messages(self, pipeline_uuid: str, session_type: str) -> list[dict]:
+        """获取消息历史"""
+        if session_type == 'person':
+            return [message.model_dump() for message in self.websocket_person_session.get_message_list(pipeline_uuid)]
+        else:
+            return [message.model_dump() for message in self.websocket_group_session.get_message_list(pipeline_uuid)]

-    def reset_session(
-        self,
-        pipeline_uuid: str,
-        session_type: str,
-        session_id: str | None = None,
-    ):
-        """Reset one pipeline/client conversation."""
-        conversation_key = self._conversation_key(pipeline_uuid, session_id)
-        session = self.websocket_person_session if session_type == 'person' else self.websocket_group_session
-        if conversation_key in session.message_lists:
-            session.message_lists[conversation_key] = []
-        if conversation_key in session.stream_message_indexes:
-            session.stream_message_indexes[conversation_key] = {}
-
-        if session_id:
-            launcher_id = (
-                f'websocketgroup_{pipeline_uuid}:{session_id}'
-                if session_type == 'group'
-                else f'websocket_{pipeline_uuid}:{session_id}'
-            )
-            self.ap.sess_mgr.session_list = [
-                candidate_session
-                for candidate_session in self.ap.sess_mgr.session_list
-                if not (
-                    str(
-                        candidate_session.launcher_type.value
-                        if hasattr(candidate_session.launcher_type, 'value')
-                        else candidate_session.launcher_type
-                    )
-                    == session_type
-                    and str(candidate_session.launcher_id) == launcher_id
-                )
-            ]
+    def reset_session(self, pipeline_uuid: str, session_type: str):
+        """重置会话"""
+        if session_type == 'person':
+            if pipeline_uuid in self.websocket_person_session.message_lists:
+                self.websocket_person_session.message_lists[pipeline_uuid] = []
+            if pipeline_uuid in self.websocket_person_session.stream_message_indexes:
+                self.websocket_person_session.stream_message_indexes[pipeline_uuid] = {}
+        else:
+            if pipeline_uuid in self.websocket_group_session.message_lists:
+                self.websocket_group_session.message_lists[pipeline_uuid] = []
+            if pipeline_uuid in self.websocket_group_session.stream_message_indexes:
+                self.websocket_group_session.stream_message_indexes[pipeline_uuid] = {}
@@ -9,16 +9,6 @@ from datetime import datetime
 import pydantic

 logger = logging.getLogger(__name__)
-_SESSION_FILTER_UNSET = object()
-
-
-def is_valid_session_id(value: str) -> bool:
-    """Accept only canonical random UUIDs for client conversation identifiers."""
-    try:
-        parsed = uuid.UUID(value)
-    except (ValueError, TypeError, AttributeError):
-        return False
-    return parsed.version == 4 and str(parsed) == value


 class WebSocketConnection(pydantic.BaseModel):
@@ -35,9 +25,6 @@ class WebSocketConnection(pydantic.BaseModel):
    session_type: str  # 'person' or 'group'
    """会话类型"""

-    session_id: str | None = None
-    """Optional client conversation identifier used by embed widgets."""
-
    websocket: typing.Any = pydantic.Field(exclude=True)
    """WebSocket连接对象 (quart.websocket)"""

@@ -78,15 +65,13 @@ class WebSocketConnectionManager:
        websocket: typing.Any,
        pipeline_uuid: str,
        session_type: str,
-        metadata: dict | None = None,
-        session_id: str | None = None,
+        metadata: dict = None,
    ) -> WebSocketConnection:
-        """Register a WebSocket connection and its optional embed session."""
+        """添加新的WebSocket连接"""
        async with self._lock:
            connection = WebSocketConnection(
                pipeline_uuid=pipeline_uuid,
                session_type=session_type,
-                session_id=session_id,
                websocket=websocket,
                metadata=metadata or {},
            )
@@ -135,25 +120,10 @@ class WebSocketConnectionManager:

            logger.debug(f'WebSocket connection disconnected: {connection_id}')

-    async def get_connection(self, connection_id: str) -> WebSocketConnection | None:
-        """Get a connection by its transport identifier."""
+    async def get_connection(self, connection_id: str) -> typing.Optional[WebSocketConnection]:
+        """获取指定连接"""
        return self.connections.get(connection_id)

-    async def get_connection_by_session_id(
-        self,
-        session_id: str,
-        pipeline_uuid: str | None = None,
-    ) -> WebSocketConnection | None:
-        """Get an active embed connection by its stable browser session identifier."""
-        for connection in self.connections.values():
-            if (
-                connection.session_id == session_id
-                and connection.is_active
-                and (pipeline_uuid is None or connection.pipeline_uuid == pipeline_uuid)
-            ):
-                return connection
-        return None
-
    async def get_connections_by_pipeline(self, pipeline_uuid: str) -> list[WebSocketConnection]:
        """获取指定流水线的所有连接"""
        connection_ids = self.pipeline_connections.get(pipeline_uuid, set())
@@ -164,30 +134,20 @@ class WebSocketConnectionManager:
        connection_ids = self.session_connections.get(session_type, set())
        return [self.connections[cid] for cid in connection_ids if cid in self.connections]

-    async def broadcast_to_pipeline(
-        self,
-        pipeline_uuid: str,
-        message: dict,
-        session_type: str | None = None,
-        session_id: typing.Any = _SESSION_FILTER_UNSET,
-    ):
-        """Broadcast a message to matching connections for one pipeline.
+    async def broadcast_to_pipeline(self, pipeline_uuid: str, message: dict, session_type: str = None):
+        """向指定流水线的所有连接广播消息

        Args:
-            pipeline_uuid: Pipeline identifier.
-            message: Serialized message to enqueue.
-            session_type: Optional session-type filter.
-            session_id: Embed conversation filter. Omit it to broadcast across
-                conversations; pass ``None`` to target non-embed connections.
+            pipeline_uuid: 流水线UUID
+            message: 要广播的消息
+            session_type: 可选的会话类型过滤器，如果提供则只向匹配的session_type连接广播
        """
        connections = await self.get_connections_by_pipeline(pipeline_uuid)

+        # 如果指定了session_type，只向匹配的连接广播
        if session_type is not None:
            connections = [conn for conn in connections if conn.session_type == session_type]

-        if session_id is not _SESSION_FILTER_UNSET:
-            connections = [conn for conn in connections if conn.session_id == session_id]
-
        tasks = []
        for conn in connections:
            tasks.append(self.send_to_connection(conn.connection_id, message))
@@ -133,9 +133,7 @@ class WecomMessageConverter(abstract_platform_adapter.AbstractMessageConverter):
        yiri_msg_list = []
        yiri_msg_list.append(platform_message.Source(id=message_id, time=datetime.datetime.now()))
        image_base64, image_format = await image.get_wecom_image_base64(pic_url=picurl)
-        yiri_msg_list.append(
-            platform_message.Image(url=picurl, base64=f'data:image/{image_format};base64,{image_base64}')
-        )
+        yiri_msg_list.append(platform_message.Image(base64=f'data:image/{image_format};base64,{image_base64}'))
        chain = platform_message.MessageChain(yiri_msg_list)

        return chain
@@ -2,7 +2,6 @@
 from __future__ import annotations

 import asyncio
-import contextlib
 import io
 import time
 import zipfile
@@ -37,12 +36,6 @@ from ..core import taskmgr
 from ..entity.persistence import plugin as persistence_plugin


-_CONNECT_TIMEOUT_SEC = 30.0
-_HEARTBEAT_INTERVAL_SEC = 20.0
-_HEARTBEAT_FAILURE_THRESHOLD = 3
-_RECONNECT_MAX_DELAY_SEC = 60.0
-
-
 class PluginRuntimeNotConnectedError(RuntimeError):
    """Raised when plugin runtime operations are requested before connection."""

@@ -77,235 +70,128 @@ class PluginRuntimeConnector(ManagedRuntimeConnector):
        super().__init__(ap)
        self.runtime_disconnect_callback = runtime_disconnect_callback
        self.is_enable_plugin = self.ap.instance_config.data.get('plugin', {}).get('enable', True)
-        self._transport_task: asyncio.Task | None = None
-        self._reconnect_task: asyncio.Task | None = None
-        self._generation = 0
-        self._connected = asyncio.Event()
-
-    def _runtime_handler(self) -> handler.RuntimeConnectionHandler:
-        runtime_handler = getattr(self, 'handler', None)
-        if runtime_handler is None:
-            raise PluginRuntimeNotConnectedError('Plugin runtime is not connected')
-        return runtime_handler
-
-    def _runtime_available(self) -> bool:
-        runtime_handler = getattr(self, 'handler', None)
-        if runtime_handler is None:
-            return False
-        # Unit-level and explicitly injected handlers don't own a transport.
-        # A managed transport must also have completed its handshake.
-        return self._transport_task is None or self._connected.is_set()

    async def heartbeat_loop(self):
-        failures = 0
-        while not self._closing:
-            await asyncio.sleep(_HEARTBEAT_INTERVAL_SEC)
+        while True:
+            await asyncio.sleep(20)
            try:
                await self.ping_plugin_runtime()
-                failures = 0
                self.ap.logger.debug('Heartbeat to plugin runtime success.')
            except Exception as e:
-                failures += 1
-                self.ap.logger.warning(
-                    f'Plugin runtime heartbeat failed ({failures}/{_HEARTBEAT_FAILURE_THRESHOLD}): {e}'
-                )
-                if failures >= _HEARTBEAT_FAILURE_THRESHOLD:
-                    self._connected.clear()
-                    self.schedule_reconnect()
-                    failures = 0
+                self.ap.logger.debug(f'Failed to heartbeat to plugin runtime: {e}')

    async def initialize(self):
        if not self.is_enable_plugin:
            self.ap.logger.info('Plugin system is disabled.')
            return

-        async with self._lifecycle_lock:
-            if self._closing:
-                raise PluginRuntimeNotConnectedError('Plugin runtime connector is shutting down')
-            if self._connected.is_set() and hasattr(self, 'handler'):
-                return
-
-            self._generation += 1
-            generation = self._generation
-            await self._stop_transport()
-            self._connected = asyncio.Event()
-            connect_errors: list[Exception] = []
-
-            async def new_connection_callback(
-                connection: base_connection.Connection,
-            ):
-                if generation != self._generation or self._closing:
-                    await connection.close()
-                    return
-                connection_ready = False
-                disconnect_notified = False
-
-                async def notify_disconnect() -> None:
-                    nonlocal disconnect_notified
-                    if (
-                        connection_ready
-                        and not disconnect_notified
-                        and generation == self._generation
-                        and not self._closing
-                    ):
-                        disconnect_notified = True
-                        self._connected.clear()
-                        await self.runtime_disconnect_callback(self)
-
-                async def disconnect_callback(
-                    rchandler: handler.RuntimeConnectionHandler,
-                ) -> bool:
-                    await notify_disconnect()
+        async def new_connection_callback(connection: base_connection.Connection):
+            async def disconnect_callback(
+                rchandler: handler.RuntimeConnectionHandler,
+            ) -> bool:
+                if platform.get_platform() == 'docker' or platform.use_websocket_to_connect_plugin_runtime():
+                    self.ap.logger.error('Disconnected from plugin runtime, trying to reconnect...')
+                    await self.runtime_disconnect_callback(self)
+                    return False
+                else:
+                    self.ap.logger.error(
+                        'Disconnected from plugin runtime, cannot automatically reconnect while LangBot connects to plugin runtime via stdio, please restart LangBot.'
+                    )
                    return False

-                runtime_handler = handler.RuntimeConnectionHandler(connection, disconnect_callback, self.ap)
-                self.handler = runtime_handler
-                self.handler_task = asyncio.create_task(runtime_handler.run())
+            self.handler = handler.RuntimeConnectionHandler(connection, disconnect_callback, self.ap)
+
+            self.handler_task = asyncio.create_task(self.handler.run())
+            _ = await self.handler.ping()
+            # Push the configured marketplace (Space) URL to the runtime so it
+            # downloads plugins from the same Space LangBot is bound to, rather
+            # than relying on the runtime's own env/default.
+            space_url = self.ap.instance_config.data.get('space', {}).get('url', '').rstrip('/')
+            if space_url:
                try:
-                    await runtime_handler.ping()
-                    space_url = self.ap.instance_config.data.get('space', {}).get('url', '').rstrip('/')
-                    if space_url:
-                        await runtime_handler.set_runtime_config(cloud_service_url=space_url)
-                    if generation == self._generation and not self._closing:
-                        connection_ready = True
-                        self._connected.set()
-                        self.ap.logger.info('Connected to plugin runtime.')
-                    await self.handler_task
-                except asyncio.CancelledError:
-                    raise
-                except Exception as exc:
-                    if not self._connected.is_set():
-                        connect_errors.append(exc)
-                        self._connected.set()
-                finally:
-                    if generation == self._generation and not self._closing:
-                        self._connected.clear()
-                        if getattr(self, 'handler', None) is runtime_handler:
-                            del self.handler
-                        await notify_disconnect()
+                    await self.handler.set_runtime_config(cloud_service_url=space_url)
+                    self.ap.logger.info(f'Pushed marketplace URL to plugin runtime: {space_url}')
+                except Exception as e:
+                    self.ap.logger.warning(f'Failed to push runtime config: {e}')
+            self.ap.logger.info('Connected to plugin runtime.')
+            await self.handler_task

-            task_coro: typing.Coroutine
-            if platform.get_platform() == 'docker' or platform.use_websocket_to_connect_plugin_runtime():
-                ws_url = self.ap.instance_config.data.get('plugin', {}).get(
-                    'runtime_ws_url',
-                    'ws://langbot_plugin_runtime:5400/control/ws',
-                )
+        task: asyncio.Task | None = None

-                async def connection_failed(ctrl, exc=None):
-                    error = exc or RuntimeError('WebSocket connection failed')
-                    connect_errors.append(error)
-                    self._connected.set()
-
-                self.ctrl = ws_client_controller.WebSocketClientController(
-                    ws_url=ws_url,
-                    make_connection_failed_callback=connection_failed,
-                )
-                task_coro = self.ctrl.run(new_connection_callback)
-            elif platform.get_platform() == 'win32':
-                await self._start_runtime_subprocess('-m', 'langbot_plugin.cli.__init__', 'rt')
-                ws_url = 'ws://localhost:5400/control/ws'
-
-                async def connection_failed(ctrl, exc=None):
-                    error = exc or RuntimeError('WebSocket connection failed')
-                    connect_errors.append(error)
-                    self._connected.set()
-
-                self.ctrl = ws_client_controller.WebSocketClientController(
-                    ws_url=ws_url,
-                    make_connection_failed_callback=connection_failed,
-                )
-                task_coro = self.ctrl.run(new_connection_callback)
-            else:
-                self.ctrl = stdio_client_controller.StdioClientController(
-                    command=sys.executable,
-                    args=['-m', 'langbot_plugin.cli.__init__', 'rt', '-s'],
-                    env=os.environ.copy(),
-                    capture_stderr=False,
-                )
-                task_coro = self.ctrl.run(new_connection_callback)
-
-            self._transport_task = asyncio.create_task(task_coro)
-            try:
-                await asyncio.wait_for(self._connected.wait(), timeout=_CONNECT_TIMEOUT_SEC)
-            except asyncio.TimeoutError as exc:
-                await self._stop_transport()
-                raise PluginRuntimeNotConnectedError('Plugin runtime did not become ready within 30 seconds') from exc
-            if connect_errors:
-                await self._stop_transport()
-                raise PluginRuntimeNotConnectedError(f'Plugin runtime connection failed: {connect_errors[-1]}')
-
-            if self.heartbeat_task is None or self.heartbeat_task.done():
-                self.heartbeat_task = asyncio.create_task(self.heartbeat_loop())
-
-    def schedule_reconnect(self) -> None:
-        if self._closing or not self.is_enable_plugin:
-            return
-        if self._reconnect_task is not None and not self._reconnect_task.done():
-            return
-        self._reconnect_task = asyncio.create_task(self._reconnect_loop())
-
-    async def _reconnect_loop(self) -> None:
-        delay = 1.0
-        try:
-            while not self._closing:
-                try:
-                    await self.initialize()
-                    return
-                except Exception as exc:
-                    self.ap.logger.warning(f'Plugin runtime reconnection failed: {exc}; retrying in {delay:.0f}s')
-                    await asyncio.sleep(delay)
-                    delay = min(delay * 2, _RECONNECT_MAX_DELAY_SEC)
-        finally:
-            self._reconnect_task = None
-
-    async def _stop_transport(self) -> None:
-        self._connected.clear()
-        runtime_handler = getattr(self, 'handler', None)
-        if runtime_handler is not None:
-            with contextlib.suppress(Exception):
-                await runtime_handler.close()
-            if getattr(self, 'handler', None) is runtime_handler:
-                del self.handler
-        tasks = [
-            task
-            for task in (
-                getattr(self, 'handler_task', None),
-                self._transport_task,
+        if platform.get_platform() == 'docker' or platform.use_websocket_to_connect_plugin_runtime():  # use websocket
+            self.ap.logger.info('use websocket to connect to plugin runtime')
+            ws_url = self.ap.instance_config.data.get('plugin', {}).get(
+                'runtime_ws_url', 'ws://langbot_plugin_runtime:5400/control/ws'
            )
-            if task is not None and task is not asyncio.current_task()
-        ]
-        for task in tasks:
-            task.cancel()
-        if tasks:
-            await asyncio.gather(*tasks, return_exceptions=True)
-        self._transport_task = None
-        if hasattr(self, 'handler_task'):
-            del self.handler_task
-        close_ctrl = getattr(getattr(self, 'ctrl', None), 'close', None)
-        if close_ctrl is not None:
-            with contextlib.suppress(Exception):
-                await close_ctrl()

-    async def aclose(self) -> None:
-        self._closing = True
-        self._generation += 1
-        reconnect_task = self._reconnect_task
-        self._reconnect_task = None
-        if reconnect_task is not None and reconnect_task is not asyncio.current_task():
-            reconnect_task.cancel()
-            await asyncio.gather(reconnect_task, return_exceptions=True)
-        if self.heartbeat_task is not None:
-            self.heartbeat_task.cancel()
-            await asyncio.gather(self.heartbeat_task, return_exceptions=True)
-            self.heartbeat_task = None
-        await self._stop_transport()
-        await self._close_managed_subprocess()
+            async def make_connection_failed_callback(
+                ctrl: ws_client_controller.WebSocketClientController,
+                exc: Exception = None,
+            ) -> None:
+                if exc is not None:
+                    self.ap.logger.error(f'Failed to connect to plugin runtime({ws_url}): {exc}')
+                else:
+                    self.ap.logger.error(f'Failed to connect to plugin runtime({ws_url}), trying to reconnect...')
+                await self.runtime_disconnect_callback(self)
+
+            self.ctrl = ws_client_controller.WebSocketClientController(
+                ws_url=ws_url,
+                make_connection_failed_callback=make_connection_failed_callback,
+            )
+            task = self.ctrl.run(new_connection_callback)
+        elif platform.get_platform() == 'win32':
+            # Due to Windows's lack of supports for both stdio and subprocess:
+            # See also: https://docs.python.org/zh-cn/3.13/library/asyncio-platforms.html
+            # We have to launch runtime via cmd but communicate via ws.
+            self.ap.logger.info('(windows) use cmd to launch plugin runtime and communicate via ws')
+
+            await self._start_runtime_subprocess('-m', 'langbot_plugin.cli.__init__', 'rt')
+
+            ws_url = 'ws://localhost:5400/control/ws'
+
+            async def make_connection_failed_callback(
+                ctrl: ws_client_controller.WebSocketClientController,
+                exc: Exception = None,
+            ) -> None:
+                if exc is not None:
+                    self.ap.logger.error(f'(windows) Failed to connect to plugin runtime({ws_url}): {exc}')
+                else:
+                    self.ap.logger.error(
+                        f'(windows) Failed to connect to plugin runtime({ws_url}), trying to reconnect...'
+                    )
+                await self.runtime_disconnect_callback(self)
+
+            self.ctrl = ws_client_controller.WebSocketClientController(
+                ws_url=ws_url,
+                make_connection_failed_callback=make_connection_failed_callback,
+            )
+            task = self.ctrl.run(new_connection_callback)
+
+        else:  # stdio
+            self.ap.logger.info('use stdio to connect to plugin runtime')
+            # cmd: lbp rt -s
+            python_path = sys.executable
+            env = os.environ.copy()
+            self.ctrl = stdio_client_controller.StdioClientController(
+                command=python_path,
+                args=['-m', 'langbot_plugin.cli.__init__', 'rt', '-s'],
+                env=env,
+            )
+            task = self.ctrl.run(new_connection_callback)
+
+        if self.heartbeat_task is None:
+            self.heartbeat_task = asyncio.create_task(self.heartbeat_loop())
+
+        asyncio.create_task(task)

    async def initialize_plugins(self):
        pass

    async def ping_plugin_runtime(self):
-        return await self._runtime_handler().ping()
+        if not hasattr(self, 'handler'):
+            raise PluginRuntimeNotConnectedError('Plugin runtime is not connected')
+
+        return await self.handler.ping()

    def _inspect_plugin_package(
        self,
@@ -587,7 +473,7 @@ class PluginRuntimeConnector(ManagedRuntimeConnector):

                                    file_bytes = download_resp.content
                                    self._inspect_plugin_package(file_bytes, task_context)
-                                    file_key = await self._runtime_handler().send_file(file_bytes, 'lbpkg')
+                                    file_key = await self.handler.send_file(file_bytes, 'lbpkg')
                                    install_info['plugin_file_key'] = file_key
                                    self.ap.logger.info(f'Transfered file {file_key} to plugin runtime')
                                    # Continue to install via runtime
@@ -610,7 +496,7 @@ class PluginRuntimeConnector(ManagedRuntimeConnector):
            plugin_author, plugin_name = self._inspect_plugin_package(file_bytes, task_context)
            if task_context is not None and plugin_author and plugin_name:
                task_context.metadata['plugin_name'] = f'{plugin_author}/{plugin_name}'
-            file_key = await self._runtime_handler().send_file(file_bytes, 'lbpkg')
+            file_key = await self.handler.send_file(file_bytes, 'lbpkg')
            install_info['plugin_file_key'] = file_key
            del install_info['plugin_file']
            self.ap.logger.info(f'Transfered file {file_key} to plugin runtime')
@@ -649,14 +535,14 @@ class PluginRuntimeConnector(ManagedRuntimeConnector):
                    plugin_author, plugin_name = self._inspect_plugin_package(file_bytes, task_context)
                    if task_context is not None and plugin_author and plugin_name:
                        task_context.metadata['plugin_name'] = f'{plugin_author}/{plugin_name}'
-                    file_key = await self._runtime_handler().send_file(file_bytes, 'lbpkg')
+                    file_key = await self.handler.send_file(file_bytes, 'lbpkg')
                    install_info['plugin_file_key'] = file_key
                    self.ap.logger.info(f'Transfered file {file_key} to plugin runtime')
            except Exception as e:
                self.ap.logger.error(f'Failed to download file from GitHub: {e}')
                raise Exception(f'Failed to download file from GitHub: {e}')

-        async for ret in self._runtime_handler().install_plugin(install_source.value, install_info):
+        async for ret in self.handler.install_plugin(install_source.value, install_info):
            current_action = ret.get('current_action', None)
            if current_action is not None:
                if task_context is not None:
@@ -680,7 +566,7 @@ class PluginRuntimeConnector(ManagedRuntimeConnector):
        plugin_name: str,
        task_context: taskmgr.TaskContext | None = None,
    ) -> dict[str, Any]:
-        async for ret in self._runtime_handler().upgrade_plugin(plugin_author, plugin_name):
+        async for ret in self.handler.upgrade_plugin(plugin_author, plugin_name):
            current_action = ret.get('current_action', None)
            if current_action is not None:
                if task_context is not None:
@@ -698,7 +584,7 @@ class PluginRuntimeConnector(ManagedRuntimeConnector):
        delete_data: bool = False,
        task_context: taskmgr.TaskContext | None = None,
    ) -> dict[str, Any]:
-        async for ret in self._runtime_handler().delete_plugin(plugin_author, plugin_name):
+        async for ret in self.handler.delete_plugin(plugin_author, plugin_name):
            current_action = ret.get('current_action', None)
            if current_action is not None:
                if task_context is not None:
@@ -713,7 +599,7 @@ class PluginRuntimeConnector(ManagedRuntimeConnector):
        if delete_data:
            if task_context is not None:
                task_context.trace('Cleaning up plugin configuration and storage...')
-            await self._runtime_handler().cleanup_plugin_data(plugin_author, plugin_name)
+            await self.handler.cleanup_plugin_data(plugin_author, plugin_name)

    async def list_plugins(self, component_kinds: list[str] | None = None) -> list[dict[str, Any]]:
        """List plugins, optionally filtered by component kinds.
@@ -724,10 +610,10 @@ class PluginRuntimeConnector(ManagedRuntimeConnector):
                           component of the specified kinds will be returned.
                           E.g., ['Command', 'EventListener', 'Tool'] for pipeline-related plugins.
        """
-        if not self.is_enable_plugin or not self._runtime_available():
+        if not self.is_enable_plugin:
            return []

-        plugins = await self._runtime_handler().list_plugins()
+        plugins = await self.handler.list_plugins()

        # Filter plugins by component kinds if specified
        if component_kinds is not None:
@@ -799,18 +685,18 @@ class PluginRuntimeConnector(ManagedRuntimeConnector):
        return plugins

    async def get_plugin_info(self, author: str, plugin_name: str) -> dict[str, Any]:
-        return await self._runtime_handler().get_plugin_info(author, plugin_name)
+        return await self.handler.get_plugin_info(author, plugin_name)

    async def set_plugin_config(self, plugin_author: str, plugin_name: str, config: dict[str, Any]) -> dict[str, Any]:
-        return await self._runtime_handler().set_plugin_config(plugin_author, plugin_name, config)
+        return await self.handler.set_plugin_config(plugin_author, plugin_name, config)

    @alru_cache(ttl=5 * 60)  # 5 minutes
    async def get_plugin_icon(self, plugin_author: str, plugin_name: str) -> dict[str, Any]:
-        return await self._runtime_handler().get_plugin_icon(plugin_author, plugin_name)
+        return await self.handler.get_plugin_icon(plugin_author, plugin_name)

    @alru_cache(ttl=5 * 60)  # 5 minutes
    async def get_plugin_readme(self, plugin_author: str, plugin_name: str, language: str = 'en') -> str:
-        return await self._runtime_handler().get_plugin_readme(plugin_author, plugin_name, language)
+        return await self.handler.get_plugin_readme(plugin_author, plugin_name, language)

    async def get_plugin_logs(
        self,
@@ -820,11 +706,11 @@ class PluginRuntimeConnector(ManagedRuntimeConnector):
        level: str | None = None,
    ) -> list[dict[str, Any]]:
        # Not cached: logs are live and change constantly.
-        return await self._runtime_handler().get_plugin_logs(plugin_author, plugin_name, limit, level)
+        return await self.handler.get_plugin_logs(plugin_author, plugin_name, limit, level)

    @alru_cache(ttl=5 * 60)
    async def get_plugin_assets(self, plugin_author: str, plugin_name: str, filepath: str) -> dict[str, Any]:
-        return await self._runtime_handler().get_plugin_assets(plugin_author, plugin_name, filepath)
+        return await self.handler.get_plugin_assets(plugin_author, plugin_name, filepath)

    async def handle_page_api(
        self,
@@ -835,15 +721,13 @@ class PluginRuntimeConnector(ManagedRuntimeConnector):
        method: str,
        body: Any = None,
    ) -> dict[str, Any]:
-        return await self._runtime_handler().handle_page_api(
-            plugin_author, plugin_name, page_id, endpoint, method, body
-        )
+        return await self.handler.handle_page_api(plugin_author, plugin_name, page_id, endpoint, method, body)

    async def get_debug_info(self) -> dict[str, Any]:
        """Get debug information including debug key and WS URL"""
-        if not self.is_enable_plugin or not self._runtime_available():
+        if not self.is_enable_plugin:
            return {}
-        return await self._runtime_handler().get_debug_info()
+        return await self.handler.get_debug_info()

    async def emit_event(
        self,
@@ -852,13 +736,13 @@ class PluginRuntimeConnector(ManagedRuntimeConnector):
    ) -> context.EventContext:
        event_ctx = context.EventContext.from_event(event)

-        if not self.is_enable_plugin or not self._runtime_available():
+        if not self.is_enable_plugin:
            event_ctx._emitted_plugins = []
            event_ctx._response_sources = []
            return event_ctx

        # Pass include_plugins to runtime for filtering
-        event_ctx_result = await self._runtime_handler().emit_event(
+        event_ctx_result = await self.handler.emit_event(
            event_ctx.model_dump(serialize_as_any=False), include_plugins=bound_plugins
        )

@@ -871,19 +755,19 @@ class PluginRuntimeConnector(ManagedRuntimeConnector):

    async def notify_plugin_diagnostic(self, diagnostic: dict[str, Any]) -> None:
        """Best-effort diagnostic forwarding to the plugin runtime."""
-        if not self.is_enable_plugin or not self._runtime_available():
+        if not self.is_enable_plugin:
            return
        try:
-            await self._runtime_handler().notify_plugin_diagnostic(diagnostic)
+            await self.handler.notify_plugin_diagnostic(diagnostic)
        except Exception as e:
            self.ap.logger.debug(f'Plugin diagnostic forwarding skipped: {e}')

    async def list_tools(self, bound_plugins: list[str] | None = None) -> list[ComponentManifest]:
-        if not self.is_enable_plugin or not self._runtime_available():
+        if not self.is_enable_plugin:
            return []

        # Pass include_plugins to runtime for filtering
-        list_tools_data = await self._runtime_handler().list_tools(include_plugins=bound_plugins)
+        list_tools_data = await self.handler.list_tools(include_plugins=bound_plugins)

        tools = [ComponentManifest.model_validate(tool) for tool in list_tools_data]

@@ -901,19 +785,16 @@ class PluginRuntimeConnector(ManagedRuntimeConnector):
            return {'error': 'Tool not found: plugin system is disabled'}

        # Pass include_plugins to runtime for validation
-        if not self._runtime_available():
-            return {'error': 'Plugin runtime is temporarily unavailable'}
-
-        return await self._runtime_handler().call_tool(
+        return await self.handler.call_tool(
            tool_name, parameters, session.model_dump(serialize_as_any=True), query_id, include_plugins=bound_plugins
        )

    async def list_commands(self, bound_plugins: list[str] | None = None) -> list[ComponentManifest]:
-        if not self.is_enable_plugin or not self._runtime_available():
+        if not self.is_enable_plugin:
            return []

        # Pass include_plugins to runtime for filtering
-        list_commands_data = await self._runtime_handler().list_commands(include_plugins=bound_plugins)
+        list_commands_data = await self.handler.list_commands(include_plugins=bound_plugins)

        commands = [ComponentManifest.model_validate(command) for command in list_commands_data]

@@ -922,15 +803,12 @@ class PluginRuntimeConnector(ManagedRuntimeConnector):
    async def execute_command(
        self, command_ctx: command_context.ExecuteContext, bound_plugins: list[str] | None = None
    ) -> typing.AsyncGenerator[command_context.CommandReturn, None]:
-        if not self.is_enable_plugin or not self._runtime_available():
+        if not self.is_enable_plugin:
            yield command_context.CommandReturn(error=command_errors.CommandNotFoundError(command_ctx.command))
            return

        # Pass include_plugins to runtime for validation
-        gen = self._runtime_handler().execute_command(
-            command_ctx.model_dump(serialize_as_any=True),
-            include_plugins=bound_plugins,
-        )
+        gen = self.handler.execute_command(command_ctx.model_dump(serialize_as_any=True), include_plugins=bound_plugins)

        async for ret in gen:
            cmd_ret = command_context.CommandReturn.model_validate(ret)
@@ -945,33 +823,27 @@ class PluginRuntimeConnector(ManagedRuntimeConnector):
        retrieval_context: dict[str, Any],
    ) -> dict[str, Any]:
        """Retrieve knowledge using a KnowledgeEngine instance."""
-        if not self.is_enable_plugin or not self._runtime_available():
+        if not self.is_enable_plugin:
            return {'results': []}

-        return await self._runtime_handler().retrieve_knowledge(
-            plugin_author, plugin_name, retriever_name, retrieval_context
-        )
+        return await self.handler.retrieve_knowledge(plugin_author, plugin_name, retriever_name, retrieval_context)

    def dispose(self):
-        """Best-effort synchronous compatibility wrapper; prefer ``aclose``."""
-        self._closing = True
+        # On non-Windows stdio mode, terminate via the controller's process handle.
+        # On Windows, the managed subprocess is cleaned up by the base class.
+        if (
+            self.is_enable_plugin
+            and hasattr(self, 'ctrl')
+            and isinstance(self.ctrl, stdio_client_controller.StdioClientController)
+        ):
+            self.ap.logger.info('Terminating plugin runtime process...')
+            self.ctrl.process.terminate()
+
+        self._dispose_subprocess()
+
        if self.heartbeat_task is not None:
            self.heartbeat_task.cancel()
            self.heartbeat_task = None
-        if self._reconnect_task is not None:
-            self._reconnect_task.cancel()
-            self._reconnect_task = None
-        for task in (
-            getattr(self, 'handler_task', None),
-            self._transport_task,
-        ):
-            if task is not None:
-                task.cancel()
-        ctrl = getattr(self, 'ctrl', None)
-        process = getattr(ctrl, 'process', None)
-        if process is not None and process.returncode is None:
-            process.terminate()
-        self._dispose_subprocess()

    @staticmethod
    def _parse_plugin_id(plugin_id: str) -> tuple[str, str]:
@@ -1001,29 +873,29 @@ class PluginRuntimeConnector(ManagedRuntimeConnector):
            context_data: IngestionContext data.
        """
        plugin_author, plugin_name = self._parse_plugin_id(plugin_id)
-        return await self._runtime_handler().rag_ingest_document(plugin_author, plugin_name, context_data)
+        return await self.handler.rag_ingest_document(plugin_author, plugin_name, context_data)

    async def call_rag_delete_document(self, plugin_id: str, document_id: str, kb_id: str) -> bool:
        plugin_author, plugin_name = self._parse_plugin_id(plugin_id)
-        return await self._runtime_handler().rag_delete_document(plugin_author, plugin_name, document_id, kb_id)
+        return await self.handler.rag_delete_document(plugin_author, plugin_name, document_id, kb_id)

    async def get_rag_creation_schema(self, plugin_id: str) -> dict[str, Any]:
        plugin_author, plugin_name = self._parse_plugin_id(plugin_id)
-        return await self._runtime_handler().get_rag_creation_schema(plugin_author, plugin_name)
+        return await self.handler.get_rag_creation_schema(plugin_author, plugin_name)

    async def get_rag_retrieval_schema(self, plugin_id: str) -> dict[str, Any]:
        plugin_author, plugin_name = self._parse_plugin_id(plugin_id)
-        return await self._runtime_handler().get_rag_retrieval_schema(plugin_author, plugin_name)
+        return await self.handler.get_rag_retrieval_schema(plugin_author, plugin_name)

    async def rag_on_kb_create(self, plugin_id: str, kb_id: str, config: dict[str, Any]) -> dict[str, Any]:
        """Notify plugin about KB creation."""
        plugin_author, plugin_name = self._parse_plugin_id(plugin_id)
-        return await self._runtime_handler().rag_on_kb_create(plugin_author, plugin_name, kb_id, config)
+        return await self.handler.rag_on_kb_create(plugin_author, plugin_name, kb_id, config)

    async def rag_on_kb_delete(self, plugin_id: str, kb_id: str) -> dict[str, Any]:
        """Notify plugin about KB deletion."""
        plugin_author, plugin_name = self._parse_plugin_id(plugin_id)
-        return await self._runtime_handler().rag_on_kb_delete(plugin_author, plugin_name, kb_id)
+        return await self.handler.rag_on_kb_delete(plugin_author, plugin_name, kb_id)

    async def call_rag_retrieve(self, plugin_id: str, retrieval_context: dict[str, Any]) -> dict[str, Any]:
        """Call plugin to retrieve knowledge.
@@ -1033,25 +905,25 @@ class PluginRuntimeConnector(ManagedRuntimeConnector):
            retrieval_context: RetrievalContext data.
        """
        plugin_author, plugin_name = self._parse_plugin_id(plugin_id)
-        return await self._runtime_handler().retrieve_knowledge(plugin_author, plugin_name, '', retrieval_context)
+        return await self.handler.retrieve_knowledge(plugin_author, plugin_name, '', retrieval_context)

    async def list_knowledge_engines(self) -> list[dict[str, Any]]:
        """List all available Knowledge Engines from plugins.

        Returns a list of Knowledge Engines with their capabilities and configuration schemas.
        """
-        if not self.is_enable_plugin or not self._runtime_available():
+        if not self.is_enable_plugin:
            return []

-        return await self._runtime_handler().list_knowledge_engines()
+        return await self.handler.list_knowledge_engines()

    async def list_parsers(self) -> list[dict[str, Any]]:
        """List all available parsers from plugins."""
-        if not self.is_enable_plugin or not self._runtime_available():
+        if not self.is_enable_plugin:
            return []
-        return await self._runtime_handler().list_parsers()
+        return await self.handler.list_parsers()

    async def call_parser(self, plugin_id: str, context_data: dict[str, Any], file_bytes: bytes) -> dict[str, Any]:
        """Call plugin to parse a document."""
        plugin_author, plugin_name = self._parse_plugin_id(plugin_id)
-        return await self._runtime_handler().parse_document(plugin_author, plugin_name, context_data, file_bytes)
+        return await self.handler.parse_document(plugin_author, plugin_name, context_data, file_bytes)
@@ -193,7 +193,6 @@ class ModelManager:
        existing_embedding_models = {
            m['uuid']: m for m in await self.ap.embedding_models_service.get_embedding_models()
        }
-        existing_rerank_models = {m['uuid']: m for m in await self.ap.rerank_models_service.get_rerank_models()}

        created = 0
        updated = 0
@@ -259,33 +258,6 @@ class ModelManager:
                        await self.ap.embedding_models_service.update_embedding_model(space_model.uuid, dict(desired))
                        updated += 1

-            elif space_model.category == 'rerank':
-                existing = existing_rerank_models.get(space_model.uuid)
-                if existing is None:
-                    await self.ap.rerank_models_service.create_rerank_model(
-                        {
-                            'uuid': space_model.uuid,
-                            'name': space_model.model_id,
-                            'provider_uuid': space_model_provider.uuid,
-                            'extra_args': {},
-                            'prefered_ranking': space_model.featured_order,
-                        },
-                        preserve_uuid=True,
-                    )
-                    created += 1
-                elif existing.get('provider_uuid') == space_model_provider.uuid:
-                    desired = {
-                        'name': space_model.model_id,
-                        'provider_uuid': space_model_provider.uuid,
-                        'prefered_ranking': space_model.featured_order,
-                    }
-                    if (
-                        existing.get('name') != desired['name']
-                        or existing.get('prefered_ranking') != desired['prefered_ranking']
-                    ):
-                        await self.ap.rerank_models_service.update_rerank_model(space_model.uuid, dict(desired))
-                        updated += 1
-
        if created or updated:
            self.ap.logger.info(f'Synced models from LangBot Space: {created} added, {updated} updated.')

@@ -13,151 +13,6 @@ import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
 import langbot_plugin.api.entities.builtin.provider.message as provider_message


-class _ThinkStripState:
-    """Stateful filter that drops think blocks across chunks."""
-
-    _THINK_OPEN = '<think>'
-    _THINK_CLOSE = '</think>'
-    _LEGACY_OPEN = 'CRETIRE_REASONING_BEGINk'
-    _LEGACY_CLOSE = 'CRETIRE_REASONING_ENDk'
-
-    def __init__(self) -> None:
-        self._pairs: tuple[tuple[str, str], ...] = (
-            (self._THINK_OPEN, self._THINK_CLOSE),
-            (self._LEGACY_OPEN, self._LEGACY_CLOSE),
-        )
-        self._open_tags = tuple(open_tag for open_tag, _close_tag in self._pairs)
-        self._buf = ''
-        self._close_tag: str | None = None
-        self._pending_initial = True
-
-    def feed(self, chunk: str) -> str:
-        """Feed a streaming delta and return user-visible content."""
-        if not chunk:
-            return chunk
-
-        text = self._buf + chunk
-        if self._close_tag is not None:
-            return self._consume_think_body(text)
-
-        return self._process_visible_text(text)
-
-    def flush(self) -> str:
-        """Release buffered visible content when the stream ends."""
-        if self._close_tag is not None:
-            self._buf = ''
-            self._close_tag = None
-            return ''
-
-        pending, self._buf = self._buf, ''
-        self._close_tag = None
-        return pending
-
-    def _consume_think_body(self, text: str) -> str:
-        close_tag = self._close_tag
-        if close_tag is None:
-            return text
-
-        close_idx = text.find(close_tag)
-        if close_idx != -1:
-            self._close_tag = None
-            self._buf = ''
-            self._pending_initial = False
-            return self._process_visible_text(text[close_idx + len(close_tag) :])
-
-        self._buf = self._close_prefix(text, close_tag)
-        return ''
-
-    def _process_visible_text(self, text: str) -> str:
-        out: list[str] = []
-        index = 0
-
-        while index < len(text):
-            if self._pending_initial:
-                open_idx, open_tag, close_tag = self._find_next_open(text, index)
-                orphan_close_idx, orphan_close_tag = self._find_next_close(text, index)
-
-                if orphan_close_idx != -1 and (open_idx == -1 or orphan_close_idx < open_idx):
-                    self._pending_initial = False
-                    index = orphan_close_idx + len(orphan_close_tag)
-                    continue
-
-                if open_idx == -1:
-                    self._buf = text[index:]
-                    return ''.join(out)
-
-                if open_idx > index:
-                    self._pending_initial = False
-                    out.append(text[index:open_idx])
-                    index = open_idx
-                    continue
-
-            open_idx, open_tag, close_tag = self._find_next_open(text, index)
-            if open_idx == -1:
-                emit_end = self._visible_emit_end(text, index)
-                out.append(text[index:emit_end])
-                if emit_end > index:
-                    self._pending_initial = False
-                self._buf = text[emit_end:]
-                return ''.join(out)
-
-            out.append(text[index:open_idx])
-            if open_idx > index:
-                self._pending_initial = False
-            body_start = open_idx + len(open_tag)
-            close_idx = text.find(close_tag, body_start)
-            if close_idx == -1:
-                self._close_tag = close_tag
-                self._buf = self._close_prefix(text[body_start:], close_tag)
-                return ''.join(out)
-
-            self._pending_initial = False
-            index = close_idx + len(close_tag)
-
-        self._buf = ''
-        return ''.join(out)
-
-    def _find_next_open(self, text: str, start: int) -> tuple[int, str, str]:
-        best_idx = -1
-        best_open = ''
-        best_close = ''
-        for open_tag, close_tag in self._pairs:
-            idx = text.find(open_tag, start)
-            if idx != -1 and (best_idx == -1 or idx < best_idx):
-                best_idx = idx
-                best_open = open_tag
-                best_close = close_tag
-        return best_idx, best_open, best_close
-
-    def _find_next_close(self, text: str, start: int) -> tuple[int, str]:
-        best_idx = -1
-        best_close = ''
-        for _open_tag, close_tag in self._pairs:
-            idx = text.find(close_tag, start)
-            if idx != -1 and (best_idx == -1 or idx < best_idx):
-                best_idx = idx
-                best_close = close_tag
-        return best_idx, best_close
-
-    def _visible_emit_end(self, text: str, start: int) -> int:
-        visible = text[start:]
-        limit = min(len(visible), max(len(open_tag) for open_tag in self._open_tags) - 1)
-        for keep in range(limit, 0, -1):
-            suffix = visible[-keep:]
-            if any(open_tag.startswith(suffix) for open_tag in self._open_tags):
-                return len(text) - keep
-        return len(text)
-
-    @staticmethod
-    def _close_prefix(text: str, close_tag: str) -> str:
-        limit = min(len(text), len(close_tag) - 1)
-        for keep in range(limit, 0, -1):
-            suffix = text[-keep:]
-            if close_tag.startswith(suffix):
-                return suffix
-        return ''
-
-
 class LiteLLMRequester(requester.ProviderAPIRequester):
    """LiteLLM unified API requester supporting chat, embedding, and rerank."""

@@ -382,25 +237,6 @@ class LiteLLMRequester(requester.ProviderAPIRequester):

        return req_messages

-    _THINK_PATTERNS: tuple[str, ...] = (
-        r'^\s*(?:(?!<think>).)*?</think>\s*',
-        r'^\s*(?:(?!CRETIRE_REASONING_BEGINk).)*?CRETIRE_REASONING_ENDk\s*',
-        r'<think>.*?</think>',
-        r'CRETIRE_REASONING_BEGINk.*?CRETIRE_REASONING_ENDk',
-    )
-
-    @classmethod
-    def _strip_think(cls, content: str) -> str:
-        """Strip chain-of-thought blocks from ``content``."""
-        if not content:
-            return content
-
-        import re
-
-        for pattern in cls._THINK_PATTERNS:
-            content = re.sub(pattern, '', content, flags=re.DOTALL)
-        return content.strip()
-
    def _process_thinking_content(self, content: str, reasoning_content: str | None, remove_think: bool) -> str:
        """Process thinking/reasoning content.

@@ -412,12 +248,20 @@ class LiteLLMRequester(requester.ProviderAPIRequester):
        Returns:
            Processed content string
        """
-        if remove_think and content:
-            content = self._strip_think(content)
+        # Extract and handle thinking tags
+        if content and 'CRETIRE_REASONING_BEGINk' in content and 'CRETIRE_REASONING_ENDk' in content:
+            import re

-        if reasoning_content and not remove_think:
-            content = f'<think>\n{reasoning_content}\n</think>\n{content or ""}'.strip()
+            think_pattern = r'CRETIRE_REASONING_BEGINk(.*?)CRETIRE_REASONING_ENDk'

+            if remove_think:
+                # Remove thinking tags and their content from output
+                content = re.sub(think_pattern, '', content, flags=re.DOTALL).strip()
+            # else: preserve thinking content as-is
+
+        # Handle separate reasoning_content field
+        # Currently we don't include reasoning_content in user-facing output regardless of remove_think
+        # because it's typically internal model reasoning, not user-visible thinking
        return content or ''

    @staticmethod
@@ -726,7 +570,6 @@ class LiteLLMRequester(requester.ProviderAPIRequester):
        chunk_idx = 0
        role = 'assistant'
        tool_call_state: dict[int, dict[str, typing.Any]] = {}
-        think_state = _ThinkStripState() if remove_think else None

        try:
            response = await acompletion(**args)
@@ -770,12 +613,6 @@ class LiteLLMRequester(requester.ProviderAPIRequester):
                        # Use reasoning_content as the displayed content
                        delta_content = reasoning_content

-                if think_state is not None and delta_content:
-                    delta_content = think_state.feed(delta_content)
-                    if not delta_content:
-                        chunk_idx += 1
-                        continue
-
                tool_calls = self._normalize_stream_tool_calls(delta.get('tool_calls'), tool_call_state)

                if chunk_idx == 0 and not delta_content and not tool_calls:
@@ -797,15 +634,6 @@ class LiteLLMRequester(requester.ProviderAPIRequester):
                yield provider_message.MessageChunk(**chunk_data)
                chunk_idx += 1

-            if think_state is not None:
-                pending_content = think_state.flush()
-                if pending_content:
-                    yield provider_message.MessageChunk(
-                        role=role,
-                        content=pending_content,
-                        is_final=True,
-                    )
-
        except Exception as e:
            self._handle_litellm_error(e)

@@ -943,21 +771,16 @@ class LiteLLMRequester(requester.ProviderAPIRequester):
        if api_key:
            headers['Authorization'] = f'Bearer {api_key}'

-        request_args = dict(extra_args)
-        rerank_url = request_args.pop('rerank_url', None)
-        rerank_path = request_args.pop('rerank_path', 'rerank')
-
        payload: dict[str, typing.Any] = {
            'model': model_name,
            'query': query,
            'documents': documents,
            'top_n': top_n,
        }
-        if request_args:
-            payload.update(request_args)
+        if extra_args:
+            payload.update(extra_args)

-        if not rerank_url:
-            rerank_url = f'{base_url}/{str(rerank_path).strip("/")}'
+        rerank_url = f'{base_url}/rerank'

        try:
            async with httpx.AsyncClient(timeout=timeout) as client:
@@ -1,5 +0,0 @@
-<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 64 64" role="img" aria-label="TokenLab">
-  <rect width="64" height="64" rx="14" fill="#111827"/>
-  <path fill="#38bdf8" d="M17 14h30v8H36v28h-8V22H17z"/>
-  <path fill="#a7f3d0" d="M40 30h8v20H28v-8h12z"/>
-</svg>
@@ -1,29 +0,0 @@
-apiVersion: v1
-kind: LLMAPIRequester
-metadata:
-  name: tokenlab-chat-completions
-  label:
-    en_US: TokenLab
-    zh_Hans: TokenLab
-  icon: tokenlab.svg
-spec:
-  litellm_provider: openai
-  config:
-  - name: base_url
-    label:
-      en_US: Base URL
-      zh_Hans: 基础 URL
-    type: string
-    required: true
-    default: https://api.tokenlab.sh/v1
-  - name: timeout
-    label:
-      en_US: Timeout
-      zh_Hans: 超时时间
-    type: integer
-    required: true
-    default: 120
-  alias: "tokenlab TokenLab gpt claude gemini deepseek qwen kimi minimax glm grok openai-compatible"
-  support_type:
-  - llm
-  provider_category: maas
@@ -49,23 +49,12 @@ def _model_has_ability(model: modelmgr_requester.RuntimeLLMModel, ability: str)
 class _StreamAccumulator:
    """Accumulate streamed content and fragmented OpenAI-style tool calls."""

-    def __init__(
-        self,
-        msg_sequence: int = 0,
-        initial_content: str | None = None,
-        remove_think: bool = False,
-    ):
+    def __init__(self, msg_sequence: int = 0, initial_content: str | None = None):
        self.tool_calls_map: dict[str, provider_message.ToolCall] = {}
        self.msg_idx = 0
        self.accumulated_content = initial_content or ''
        self.last_role = 'assistant'
        self.msg_sequence = msg_sequence
-        self.remove_think = remove_think
-        self._think_state = None
-        if remove_think:
-            from ..modelmgr.requesters.litellmchat import _ThinkStripState
-
-            self._think_state = _ThinkStripState()

    def add(self, msg: provider_message.MessageChunk) -> provider_message.MessageChunk | None:
        self.msg_idx += 1
@@ -74,10 +63,7 @@ class _StreamAccumulator:
            self.last_role = msg.role

        if msg.content:
-            content = msg.content
-            if self._think_state is not None:
-                content = self._think_state.feed(content)
-            self.accumulated_content += content
+            self.accumulated_content += msg.content

        if msg.tool_calls:
            for tool_call in msg.tool_calls:
@@ -93,14 +79,11 @@ class _StreamAccumulator:
                if tool_call.function and tool_call.function.arguments:
                    self.tool_calls_map[tool_call.id].function.arguments += tool_call.function.arguments

-        if msg.is_final:
-            self._flush_think_state()
-
        if self.msg_idx % 8 == 0 or msg.is_final:
            self.msg_sequence += 1
            return provider_message.MessageChunk(
                role=self.last_role,
-                content=self._maybe_strip_think(self.accumulated_content),
+                content=self.accumulated_content,
                tool_calls=list(self.tool_calls_map.values()) if (self.tool_calls_map and msg.is_final) else None,
                is_final=msg.is_final,
                msg_sequence=self.msg_sequence,
@@ -109,29 +92,13 @@ class _StreamAccumulator:
        return None

    def final_message(self) -> provider_message.MessageChunk:
-        self._flush_think_state()
        return provider_message.MessageChunk(
            role=self.last_role,
-            content=self._maybe_strip_think(self.accumulated_content),
+            content=self.accumulated_content,
            tool_calls=list(self.tool_calls_map.values()) if self.tool_calls_map else None,
            msg_sequence=self.msg_sequence,
        )

-    def _maybe_strip_think(self, content: str) -> str:
-        if not self.remove_think or not content:
-            return content
-
-        from ..modelmgr.requesters.litellmchat import LiteLLMRequester
-
-        return LiteLLMRequester._strip_think(content)
-
-    def _flush_think_state(self) -> None:
-        if self._think_state is None:
-            return
-        pending = self._think_state.flush()
-        if pending:
-            self.accumulated_content += pending
-

@runner.runner_class('local-agent')
 class LocalAgentRunner(runner.RequestRunner):
@@ -481,7 +448,7 @@ class LocalAgentRunner(runner.RequestRunner):
        except AttributeError:
            is_stream = False

-        remove_think = ((query.pipeline_config.get('output') or {}).get('misc') or {}).get('remove-think', False)
+        remove_think = query.pipeline_config['output'].get('misc', '').get('remove-think')

        # Build ordered candidate list (primary + fallbacks)
        candidates = await self._get_model_candidates(query)
@@ -505,7 +472,7 @@ class LocalAgentRunner(runner.RequestRunner):
            final_msg = msg
        else:
            # Streaming: invoke with fallback
-            stream_accumulator = _StreamAccumulator(msg_sequence=1, remove_think=remove_think)
+            stream_accumulator = _StreamAccumulator(msg_sequence=1)

            stream_src, use_llm_model = await self._invoke_stream_with_fallback(
                query,
@@ -523,6 +490,7 @@ class LocalAgentRunner(runner.RequestRunner):
            final_msg = stream_accumulator.final_message()

        pending_tool_calls = final_msg.tool_calls
+        first_content = final_msg.content
        if isinstance(final_msg, provider_message.MessageChunk):
            first_end_sequence = final_msg.msg_sequence

@@ -605,14 +573,9 @@ class LocalAgentRunner(runner.RequestRunner):
            )

            if is_stream:
-                # Do NOT re-seed the accumulator with first_content:
-                # the previous round's text was already pushed to the
-                # platform adapter. Re-seeding would cause every
-                # subsequent round to repeat the entire opening line,
-                # which the platform then forwards as a duplicate message.
                stream_accumulator = _StreamAccumulator(
                    msg_sequence=first_end_sequence,
-                    remove_think=remove_think,
+                    initial_content=first_content,
                )

                tool_stream_src = use_llm_model.provider.invoke_llm_stream(
@@ -3,17 +3,14 @@ from __future__ import annotations
 import base64
 import enum
 import json
-import math
 import re
 import time
 import typing
 from contextlib import AsyncExitStack, asynccontextmanager
-from datetime import timedelta
 import traceback
 from langbot_plugin.api.entities.events import pipeline_query
 import sqlalchemy
 import asyncio
-import hashlib
 import httpx

 import uuid as uuid_module
@@ -29,13 +26,7 @@ from ....core import app
 import langbot_plugin.api.entities.builtin.resource.tool as resource_tool
 import langbot_plugin.api.entities.builtin.provider.message as provider_message
 from ....entity.persistence import mcp as persistence_mcp
-from .mcp_stdio import (
-    BoxStdioSessionRuntime,
-    MCPServerBoxConfig as MCPServerBoxConfig,  # noqa: F401 - public re-export
-    MCPSessionErrorPhase,
-    _ColdStartRetry,
-    _get_default_memory_mb,
-)  # noqa: F401
+from .mcp_stdio import BoxStdioSessionRuntime, MCPServerBoxConfig, MCPSessionErrorPhase, _ColdStartRetry  # noqa: F401

 # Synthesized LLM tools for MCP resources (not from server tools/list).
 # Dispatched in MCPLoader.invoke_tool; placeholder func on LLMTool is never used.
@@ -53,7 +44,6 @@ MCP_RESOURCE_CONTEXT_MAX_BYTES = 96 * 1024
 MCP_RESOURCE_TRACE_QUERY_KEY = '_mcp_resource_reads'
 MCP_RESOURCE_LINKS_QUERY_KEY = '_mcp_resource_links'
 MCP_RESOURCE_CONTEXT_QUERY_KEY = '_mcp_resource_context'
-MCP_TOOL_CALL_TIMEOUT_DEFAULT_SECONDS = 300.0

 TEXT_LIKE_MIME_TYPES = {
    'application/json',
@@ -206,20 +196,6 @@ class _TransportReconnect(Exception):
    """


-class _CallerReconnect(Exception):
-    """Internal signal: a tool/resource call hit a server-expired session
-    (e.g. a UDP/HTTP MCP server's own session timeout) and asked the owning
-    lifecycle loop to rebuild the connection.
-
-    Like _TransportReconnect, this does NOT consume the fatal retry budget —
-    the server-side timeout is expected, recurring behavior, not a failure.
-    """
-
-
-class MCPToolCallTimeoutError(TimeoutError):
-    """An MCP tool call exceeded its configured per-server deadline."""
-
-
 class RuntimeMCPSession:
    """运行时 MCP 会话"""

@@ -269,9 +245,6 @@ class RuntimeMCPSession:
        self.ap = ap
        self.enable = enable
        self.session = None
-        self.tool_call_timeout_sec = self._parse_tool_call_timeout(
-            server_config.get('tool_call_timeout_sec', MCP_TOOL_CALL_TIMEOUT_DEFAULT_SECONDS)
-        )

        # Transient test sessions (created from the config page "test" button,
        # which carry no persisted server UUID) must NOT share the live
@@ -292,12 +265,6 @@ class RuntimeMCPSession:
        self._lifecycle_task = None
        self._shutdown_event = asyncio.Event()
        self._ready_event = asyncio.Event()
-        # Signaled by callers (invoke_mcp_tool / read_resource_envelope) when
-        # they see a server-expired session; the lifecycle loop reconnects and
-        # sets _reconnected_event so all callers waiting on this cycle resume
-        # together, instead of each racing to rebuild the session itself.
-        self._reconnect_event = asyncio.Event()
-        self._reconnected_event: asyncio.Event | None = None
        # Set transiently when a WS transport drop should NOT stop the managed
        # process (it will be re-attached on the next initialize()).
        self._preserve_managed_process = False
@@ -312,39 +279,21 @@ class RuntimeMCPSession:
        self._box_stdio_runtime = BoxStdioSessionRuntime(self)
        self.box_config = self._box_stdio_runtime.config

-    def _parse_tool_call_timeout(self, value: typing.Any) -> float:
-        """Return a safe tool-call timeout; zero explicitly disables it."""
-        try:
-            timeout = -1 if isinstance(value, bool) else float(value)
-            if timeout > 0:
-                # Validate the exact conversion used for each call here, so a
-                # finite-but-enormous manual config cannot fail at invocation.
-                timedelta(seconds=timeout)
-        except (TypeError, ValueError, OverflowError):
-            timeout = -1
-
-        if not math.isfinite(timeout) or timeout < 0:
-            self.ap.logger.warning(
-                f'Invalid MCP tool call timeout {value!r} for {self.server_name}; '
-                f'using {MCP_TOOL_CALL_TIMEOUT_DEFAULT_SECONDS:g} seconds'
-            )
-            return MCP_TOOL_CALL_TIMEOUT_DEFAULT_SECONDS
-        return timeout
-
    async def _init_stdio_python_server(self):
        if self._uses_box_stdio():
            await self._box_stdio_runtime.initialize()
            return

-        # Box is configured but explicitly disabled. Refuse stdio MCP rather
-        # than silently falling through to host-stdio — the operator asked for
-        # the sandbox and the failure mode should be visible. An enabled Box
-        # that is reconnecting is handled above and waits for availability.
+        # Box is configured (ap.box_service exists) but currently unavailable
+        # (disabled by config or connection failed). Refuse stdio MCP rather
+        # than silently falling through to host-stdio — the operator asked
+        # for the sandbox and the failure mode should be visible.
        #
        # Set ``error_phase = BOX_UNAVAILABLE`` BEFORE raising so the retry
-        # wrapper can distinguish a deliberately disabled Box from an enabled
-        # runtime that is still reconnecting. Keep the message itself short —
-        # the frontend ignores it for this phase.
+        # wrapper can short-circuit (retrying is pointless when Box is
+        # deliberately off) and the frontend can render a localized,
+        # actionable message instead of this raw RuntimeError. Keep the
+        # message itself short — the frontend ignores it for this phase.
        box_service = getattr(self.ap, 'box_service', None)
        if box_service is not None and not getattr(box_service, 'available', False):
            self.error_phase = MCPSessionErrorPhase.BOX_UNAVAILABLE
@@ -480,19 +429,12 @@ class RuntimeMCPSession:
            if self._uses_box_stdio():
                monitor_task = asyncio.create_task(self._box_stdio_runtime.monitor_process_health())
                shutdown_task = asyncio.create_task(self._shutdown_event.wait())
-                reconnect_task = asyncio.create_task(self._reconnect_event.wait())
                done, pending = await asyncio.wait(
-                    [shutdown_task, monitor_task, reconnect_task],
+                    [shutdown_task, monitor_task],
                    return_when=asyncio.FIRST_COMPLETED,
                )
                for task in pending:
                    task.cancel()
-                if reconnect_task in done and not self._shutdown_event.is_set():
-                    self._reconnect_event.clear()
-                    self.ap.logger.info(
-                        f'MCP session {self.server_name}: caller requested reconnect (server session expired)'
-                    )
-                    raise _CallerReconnect('Caller requested reconnect after session expiry')
                for task in done:
                    if task is monitor_task and not self._shutdown_event.is_set():
                        # The monitor completed. This is EITHER the managed
@@ -520,20 +462,7 @@ class RuntimeMCPSession:
                        self.error_phase = MCPSessionErrorPhase.RUNTIME
                        raise Exception('Box managed process exited unexpectedly')
            else:
-                shutdown_task = asyncio.create_task(self._shutdown_event.wait())
-                reconnect_task = asyncio.create_task(self._reconnect_event.wait())
-                done, pending = await asyncio.wait(
-                    [shutdown_task, reconnect_task],
-                    return_when=asyncio.FIRST_COMPLETED,
-                )
-                for task in pending:
-                    task.cancel()
-                if reconnect_task in done and not self._shutdown_event.is_set():
-                    self._reconnect_event.clear()
-                    self.ap.logger.info(
-                        f'MCP session {self.server_name}: caller requested reconnect (server session expired)'
-                    )
-                    raise _CallerReconnect('Caller requested reconnect after session expiry')
+                await self._shutdown_event.wait()

        except _ColdStartRetry:
            # Cold-start in progress: set the preserve flag BEFORE the finally
@@ -592,43 +521,6 @@ class RuntimeMCPSession:
                self.error_phase = None
                await asyncio.sleep(1)
                continue
-            except _CallerReconnect:
-                # A tool/resource call hit a server-expired session and asked us
-                # to rebuild the connection. Same treatment as _TransportReconnect:
-                # reconnect immediately WITHOUT consuming the fatal retry budget.
-                # Wake any callers waiting in _trigger_reconnect() regardless of
-                # outcome so they don't block for the full timeout.
-                reconnected_event = self._reconnected_event
-                if self._shutdown_event.is_set():
-                    if reconnected_event is not None:
-                        reconnected_event.set()
-                    return
-                self.status = MCPSessionStatus.CONNECTING
-                self.error_message = None
-                self.error_phase = None
-                try:
-                    if self.server_config['mode'] == 'stdio':
-                        await self._init_stdio_python_server()
-                    elif self.server_config['mode'] == 'remote':
-                        await self._init_remote_server()
-                    elif self.server_config['mode'] == 'sse':
-                        await self._init_sse_server()
-                    elif self.server_config['mode'] == 'http':
-                        await self._init_streamable_http_server()
-                    await self.refresh()
-                    self.status = MCPSessionStatus.CONNECTED
-                    self.ap.logger.info(f'MCP session {self.server_name} reconnected successfully after session expiry')
-                except Exception as reconnect_err:
-                    self.status = MCPSessionStatus.ERROR
-                    self.error_message = str(reconnect_err)
-                    self.ap.logger.error(
-                        f'MCP session {self.server_name}: reconnect after session expiry failed: '
-                        f'{self._describe_exception(reconnect_err)}'
-                    )
-                finally:
-                    if reconnected_event is not None:
-                        reconnected_event.set()
-                continue
            except _ColdStartRetry as e:
                # The managed process is alive but still cold-starting (e.g.
                # `npx -y <pkg>` is still installing) and cannot yet answer the
@@ -648,29 +540,17 @@ class RuntimeMCPSession:
                await asyncio.sleep(2)
                continue
            except Exception as e:
+                self.retry_count = attempt + 1
                if self._shutdown_event.is_set():
                    return  # Shutdown requested, don't retry
+                # BOX_UNAVAILABLE is a deliberate refusal, not a transient
+                # failure — retrying produces log spam and a misleading
+                # "Failed after N attempts" message. Surface it immediately.
                if self.error_phase == MCPSessionErrorPhase.BOX_UNAVAILABLE:
-                    box_service = getattr(self.ap, 'box_service', None)
-                    if box_service is not None and getattr(box_service, 'enabled', True):
-                        # Box is configured and may recover independently of
-                        # this MCP session. Keep retrying without consuming the
-                        # fatal budget; _wait_for_box_runtime() rate-limits the
-                        # loop to one warning per startup timeout.
-                        self.status = MCPSessionStatus.CONNECTING
-                        self.error_message = None
-                        self.error_phase = None
-                        await asyncio.sleep(1)
-                        continue
-                    # Explicitly disabled Box is a deliberate refusal, not a
-                    # transient failure. Surface it immediately without log
-                    # spam or a misleading "Failed after N attempts" message.
-                    self.retry_count = attempt + 1
                    self.status = MCPSessionStatus.ERROR
                    self.error_message = str(e)
                    self._ready_event.set()
                    return
-                self.retry_count = attempt + 1
                if attempt >= self._MAX_RETRIES:
                    self.status = MCPSessionStatus.ERROR
                    self.error_message = f'Failed after {self._MAX_RETRIES + 1} attempts: {self._describe_exception(e)}'
@@ -743,47 +623,6 @@ class RuntimeMCPSession:
                return True
        return False

-    @staticmethod
-    def _is_session_terminated(exc: BaseException) -> bool:
-        """Whether exc indicates the server-side session expired.
-
-        Long-lived MCP servers (notably UDP/HTTP transports) commonly enforce
-        their own session timeout (e.g. ~30 minutes); once it fires, any call
-        on the cached session raises this rather than a transport-level error.
-        """
-        for leaf in RuntimeMCPSession._iter_exception_leaves(exc):
-            msg = str(leaf).lower()
-            if 'session terminated' in msg or 'session expired' in msg:
-                return True
-        return False
-
-    _RECONNECT_WAIT_TIMEOUT = 30.0
-
-    async def _trigger_reconnect(self) -> bool:
-        """Ask the owning lifecycle loop to rebuild the session and wait for it.
-
-        Concurrent callers that hit the timeout at the same time all signal
-        the same _reconnect_event and await the same _reconnected_event, so
-        the lifecycle loop reconnects once and every caller resumes together
-        — instead of each caller racing to rebuild the session itself.
-
-        Returns True if reconnection succeeded within the timeout.
-        """
-        if self._shutdown_event.is_set():
-            return False
-
-        if self._reconnected_event is None or self._reconnected_event.is_set():
-            self._reconnected_event = asyncio.Event()
-        reconnected_event = self._reconnected_event
-        self._reconnect_event.set()
-
-        try:
-            await asyncio.wait_for(reconnected_event.wait(), timeout=self._RECONNECT_WAIT_TIMEOUT)
-            return self.status == MCPSessionStatus.CONNECTED
-        except asyncio.TimeoutError:
-            self.ap.logger.warning(f'MCP session {self.server_name} reconnect timed out')
-            return False
-
    _MONITOR_POLL_INTERVAL = 5
    _MONITOR_MAX_CONSECUTIVE_ERRORS = 3

@@ -992,58 +831,21 @@ class RuntimeMCPSession:
        arguments: dict,
        query: pipeline_query.Query | None = None,
    ) -> list[provider_message.ContentElement]:
-        for attempt in range(2):
-            if not self.session:
-                raise Exception('MCP session is not connected')
+        if not self.session:
+            raise Exception('MCP session is not connected')

-            try:
-                read_timeout = timedelta(seconds=self.tool_call_timeout_sec) if self.tool_call_timeout_sec > 0 else None
-                result = await self.session.call_tool(
-                    tool_name,
-                    arguments,
-                    read_timeout_seconds=read_timeout,
-                )
-            except Exception as e:
-                if self._is_tool_call_timeout(e):
-                    self.ap.logger.warning(
-                        f'MCP tool {tool_name} on {self.server_name} timed out after '
-                        f'{self.tool_call_timeout_sec:g} seconds'
-                    )
-                    raise MCPToolCallTimeoutError(
-                        f"MCP tool '{tool_name}' on server '{self.server_name}' timed out after "
-                        f'{self.tool_call_timeout_sec:g} seconds'
-                    ) from e
-                if attempt == 0 and self._is_session_terminated(e):
-                    self.ap.logger.warning(
-                        f'MCP tool {tool_name} on {self.server_name} got session terminated, triggering reconnect...'
-                    )
-                    if await self._trigger_reconnect():
-                        continue
-                raise
-
-            if result.isError:
-                error_texts = []
-                for content in result.content:
-                    if getattr(content, 'type', '') == 'text':
-                        error_texts.append(content.text)
-                raise Exception('\n'.join(error_texts) if error_texts else 'Unknown error from MCP tool')
-
-            result_contents: list[provider_message.ContentElement] = []
+        result = await self.session.call_tool(tool_name, arguments)
+        if result.isError:
+            error_texts = []
            for content in result.content:
-                result_contents.extend(self._content_to_provider_elements(content, query=query, source_tool=tool_name))
-            return result_contents
+                if getattr(content, 'type', '') == 'text':
+                    error_texts.append(content.text)
+            raise Exception('\n'.join(error_texts) if error_texts else 'Unknown error from MCP tool')

-        raise Exception('MCP session is not connected')
-
-    @staticmethod
-    def _is_tool_call_timeout(exc: BaseException) -> bool:
-        """Recognize the MCP SDK's per-request timeout without retrying it."""
-        return any(
-            isinstance(leaf, McpError)
-            and leaf.error.code == httpx.codes.REQUEST_TIMEOUT
-            and leaf.error.message.startswith('Timed out while waiting for response')
-            for leaf in RuntimeMCPSession._iter_exception_leaves(exc)
-        )
+        result_contents: list[provider_message.ContentElement] = []
+        for content in result.content:
+            result_contents.extend(self._content_to_provider_elements(content, query=query, source_tool=tool_name))
+        return result_contents

    def get_tools(self) -> list[resource_tool.LLMTool]:
        return self.functions
@@ -1108,21 +910,7 @@ class RuntimeMCPSession:
            self._record_resource_read_trace(query, envelope)
            return envelope

-        result = None
-        for attempt in range(2):
-            if not self.session:
-                raise Exception('MCP session is not connected')
-            try:
-                result = await self.session.read_resource(AnyUrl(uri))
-                break
-            except Exception as e:
-                if attempt == 0 and self._is_session_terminated(e):
-                    self.ap.logger.warning(
-                        f'MCP resource read on {self.server_name} got session terminated, triggering reconnect...'
-                    )
-                    if await self._trigger_reconnect():
-                        continue
-                raise
+        result = await self.session.read_resource(AnyUrl(uri))
        contents: list[dict] = []
        total_bytes = 0
        truncated_any = False
@@ -1277,34 +1065,15 @@ class RuntimeMCPSession:
        return self._box_stdio_runtime.uses_box_stdio()

    def _build_box_session_id(self) -> str:
-        # Compatible MCP servers share a session and remain isolated by
-        # process_id. A server with a different immutable resource profile gets
-        # another session; Docker/E2B cannot change memory/image/etc. after a
-        # session has been created.
-        config = self._box_stdio_runtime.config
-        default_memory = _get_default_memory_mb(self.ap)
-        profile = {
-            'image': config.image,
-            'network': config.network,
-            'host_path_mode': config.host_path_mode,
-            'cpus': config.cpus,
-            'memory_mb': config.memory_mb or default_memory,
-            'pids_limit': config.pids_limit,
-            'read_only_rootfs': (config.read_only_rootfs if config.read_only_rootfs is not None else False),
-        }
-        default_profile = {
-            'image': None,
-            'network': 'on',
-            'host_path_mode': 'ro',
-            'cpus': None,
-            'memory_mb': default_memory,
-            'pids_limit': None,
-            'read_only_rootfs': False,
-        }
-        if profile == default_profile:
-            return 'mcp-shared'
-        digest = hashlib.sha256(json.dumps(profile, sort_keys=True).encode('utf-8')).hexdigest()[:12]
-        return f'mcp-shared-{digest}'
+        # Both live servers and transient config-page tests share ONE Box
+        # session ('mcp-shared'). A test therefore reuses the already-running
+        # container (and, for an existing server, its live managed process)
+        # instead of paying a full per-test session cold-start + dependency
+        # bootstrap. Isolation between a test and the live servers is provided
+        # at the *process* level: each server/test has its own process_id and a
+        # test only ever stops its own process_id (see cleanup_session), so it
+        # never disturbs another server's process or the shared session itself.
+        return 'mcp-shared'

    def _rewrite_path(self, path: str, host_path: str | None) -> str:
        return self._box_stdio_runtime.rewrite_path(path, host_path)
@@ -1887,23 +1656,11 @@ class MCPLoader(loader.ToolLoader):
    async def shutdown(self):
        """关闭所有工具"""
        self.ap.logger.info('Shutting down all MCP sessions...')
-
-        hosted_tasks = [task for task in self._hosted_mcp_tasks if not task.done()]
-        for task in hosted_tasks:
-            task.cancel()
-        if hosted_tasks:
-            await asyncio.gather(*hosted_tasks, return_exceptions=True)
-        self._hosted_mcp_tasks.clear()
-
-        async def shutdown_session(server_name: str, session: RuntimeMCPSession) -> None:
+        for server_name, session in list(self.sessions.items()):
            try:
                await session.shutdown()
                self.ap.logger.debug(f'Shutdown MCP session: {server_name}')
            except Exception as e:
                self.ap.logger.error(f'Error shutting down MCP session {server_name}: {e}\n{traceback.format_exc()}')
-
-        await asyncio.gather(
-            *(shutdown_session(server_name, session) for server_name, session in list(self.sessions.items()))
-        )
        self.sessions.clear()
        self.ap.logger.info('All MCP sessions shutdown complete')
@@ -185,10 +185,11 @@ class BoxStdioSessionRuntime:
        box_service = getattr(self.ap, 'box_service', None)
        if box_service is None:
            return False
-        # An enabled Box service remains the required transport while it is
-        # reconnecting. initialize() waits for availability instead of
-        # permanently failing the MCP server or falling through to host stdio.
-        return bool(getattr(box_service, 'enabled', True))
+        # When Box is configured but currently unavailable (disabled or
+        # connection failed), do NOT silently fall through to host-stdio —
+        # that would bypass the sandbox the operator asked for. The caller
+        # is expected to refuse the stdio MCP server with a clear error.
+        return bool(getattr(box_service, 'available', False))

    async def initialize(self) -> None:
        await self._wait_for_box_runtime()
@@ -487,7 +488,7 @@ class BoxStdioSessionRuntime:
                )
                warned = True
            if asyncio.get_running_loop().time() >= deadline:
-                self.owner.error_phase = MCPSessionErrorPhase.BOX_UNAVAILABLE
+                self.owner.error_phase = MCPSessionErrorPhase.SESSION_CREATE
                raise Exception(f'Box runtime is not available after {int(timeout_sec)} seconds')
            await asyncio.sleep(1)

@@ -57,7 +57,7 @@ class NativeToolLoader(loader.ToolLoader):
        return await is_box_backend_available(self.ap)

    async def get_tools(self, bound_plugins: list[str] | None = None) -> list[resource_tool.LLMTool]:
-        if not await self._is_sandbox_available():
+        if not self._is_sandbox_available():
            return []
        if self._tools is None:
            self._tools = [
@@ -71,7 +71,7 @@ class NativeToolLoader(loader.ToolLoader):
        return list(self._tools)

    async def has_tool(self, name: str) -> bool:
-        return name in _ALL_TOOL_NAMES and await self._is_sandbox_available()
+        return name in _ALL_TOOL_NAMES and self._is_sandbox_available()

    async def invoke_tool(self, name: str, parameters: dict, query: pipeline_query.Query):
        if name == EXEC_TOOL_NAME:
@@ -652,9 +652,12 @@ else:
        if callable(refresh_skill):
            refresh_skill(selected_skill.get('name', ''))

-    async def _is_sandbox_available(self) -> bool:
-        """Refresh backend availability so Box reconnects restore tool exposure."""
-        self._backend_available = await self._check_backend_available()
+    def _is_sandbox_available(self) -> bool:
+        """Check if sandbox backend is available.
+
+        This checks the cached backend availability from initialization,
+        not just whether the box_service process is running.
+        """
        return bool(self._backend_available)

    def _build_exec_tool(self) -> resource_tool.LLMTool:
@@ -49,27 +49,19 @@ class SkillToolLoader(loader.ToolLoader):
        return await is_box_backend_available(self.ap)

    async def get_tools(self, bound_plugins: list[str] | None = None) -> list[resource_tool.LLMTool]:
-        if not await self._is_available():
+        if not self._is_available():
            return []
-        if not self._tools:
-            self._tools = [
-                self._build_activate_skill_tool(),
-                self._build_register_skill_tool(),
-            ]
        return list(self._tools)

    async def has_tool(self, name: str) -> bool:
-        return await self._is_available() and name in SKILL_TOOL_NAMES
+        return self._is_available() and name in SKILL_TOOL_NAMES

-    async def _is_available(self) -> bool:
+    def _is_available(self) -> bool:
        """Check if skill tools should be available.

        Skill tools require both a skill manager and a sandbox backend.
        """
-        if not self._has_skill_manager():
-            return False
-        self._sandbox_available = await self._check_sandbox_available()
-        return self._sandbox_available
+        return self._has_skill_manager() and self._sandbox_available

    async def invoke_tool(self, name: str, parameters: dict, query) -> typing.Any:
        if name == ACTIVATE_SKILL_TOOL_NAME:
@@ -3,7 +3,6 @@
 from __future__ import annotations

 import asyncio
-import contextlib
 import os
 import sys
 from typing import TYPE_CHECKING, Awaitable, Callable
@@ -28,8 +27,6 @@ class ManagedRuntimeConnector:
        self.ap = ap
        self.runtime_subprocess = None
        self.runtime_subprocess_task = None
-        self._lifecycle_lock = asyncio.Lock()
-        self._closing = False

    async def _start_runtime_subprocess(self, *args: str) -> None:
        """Launch a local runtime as a subprocess of the current Python interpreter.
@@ -89,23 +86,3 @@ class ManagedRuntimeConnector:
        if self.runtime_subprocess_task is not None:
            self.runtime_subprocess_task.cancel()
            self.runtime_subprocess_task = None
-
-    async def _close_managed_subprocess(self) -> None:
-        """Terminate, escalate, and reap the optional owned subprocess."""
-        process = self.runtime_subprocess
-        wait_task = self.runtime_subprocess_task
-        if process is not None and process.returncode is None:
-            with contextlib.suppress(ProcessLookupError):
-                process.terminate()
-            try:
-                await asyncio.wait_for(process.wait(), timeout=3)
-            except asyncio.TimeoutError:
-                with contextlib.suppress(ProcessLookupError):
-                    process.kill()
-                await process.wait()
-        if wait_task is not None and wait_task is not asyncio.current_task():
-            if not wait_task.done():
-                wait_task.cancel()
-            await asyncio.gather(wait_task, return_exceptions=True)
-        self.runtime_subprocess = None
-        self.runtime_subprocess_task = None
@@ -119,8 +119,7 @@ class ValkeySearchVectorDatabase(VectorDatabase):
    def __init__(self, ap: app.Application):
        if not VALKEY_SEARCH_AVAILABLE:
            raise ImportError(
-                'valkey-glide is not installed or is unavailable on this platform. '
-                "On Linux or macOS, install it with: pip install 'valkey-glide>=2.4.1,<3.0.0'"
+                "valkey-glide is not installed. Install it with: pip install 'valkey-glide>=2.4.1,<3.0.0'"
            )

        self.ap = ap
@@ -143,11 +143,6 @@ box:
    backend: 'local'  # 'local' (Docker/nsjail), 'docker', 'nsjail', or 'e2b'. Can be written via BOX__BACKEND.
    runtime:
        endpoint: ''  # External Box Runtime base URL, e.g. 'ws://127.0.0.1:5410'. Leave empty for local auto-managed runtime.
-    limits:
-        max_sessions: 64  # Includes persistent sessions. New sessions fail explicitly when this cap is reached.
-        max_managed_processes: 64  # Maximum concurrently running stdio MCP / managed processes.
-        max_completed_processes: 256  # Global cap for retained exited-process diagnostics.
-        completed_process_retention_sec: 300  # Keep exited-process diagnostics before releasing memory.
    local:
        profile: 'default'
        image: ''  # Custom local sandbox image. Leave empty to use the profile default.
@@ -303,60 +303,11 @@
    '<svg viewBox="0 0 24 24"><path d="M14 2H6a2 2 0 00-2 2v16a2 2 0 002 2h12a2 2 0 002-2V8l-6-6zM6 20V4h7v5h5v11H6z"/><path d="M8 17l2.5-3.5L13 17l2-2.5L18 17H8z"/></svg>';

  // ========== State ==========
-  function createSessionId() {
-    if (window.crypto && typeof window.crypto.randomUUID === "function") {
-      return window.crypto.randomUUID();
-    }
-
-    var bytes = new Uint8Array(16);
-    if (!window.crypto || typeof window.crypto.getRandomValues !== "function") {
-      throw new Error("Secure random number generation is unavailable");
-    }
-    window.crypto.getRandomValues(bytes);
-    bytes[6] = (bytes[6] & 15) | 64;
-    bytes[8] = (bytes[8] & 63) | 128;
-    var hex = Array.prototype.map
-      .call(bytes, function (value) {
-        return value.toString(16).padStart(2, "0");
-      })
-      .join("");
-    return (
-      hex.slice(0, 8) +
-      "-" +
-      hex.slice(8, 12) +
-      "-" +
-      hex.slice(12, 16) +
-      "-" +
-      hex.slice(16, 20) +
-      "-" +
-      hex.slice(20)
-    );
-  }
-
-  function getOrCreateSessionId() {
-    var storageKey = "langbot_embed_session_" + CONFIG.botUuid;
-    try {
-      var stored = window.sessionStorage.getItem(storageKey);
-      if (
-        /^[a-f0-9]{8}-[a-f0-9]{4}-4[a-f0-9]{3}-[89ab][a-f0-9]{3}-[a-f0-9]{12}$/.test(
-          stored || "",
-        )
-      )
-        return stored;
-      var created = createSessionId();
-      window.sessionStorage.setItem(storageKey, created);
-      return created;
-    } catch (e) {
-      return createSessionId();
-    }
-  }
-
  var state = {
    isOpen: false,
    isConnected: false,
    ws: null,
    connectionId: null,
-    sessionId: getOrCreateSessionId(),
    reconnectAttempts: 0,
    heartbeatTimer: null,
    messages: [],
@@ -364,9 +315,6 @@
    isStreaming: false,
    streamingMsgId: null,
    historyLoaded: false,
-    hasConnected: false,
-    messageVersion: 0,
-    historyReloadTimer: null,
    pendingImage: null,
    feedbackState: {},
  };
@@ -525,9 +473,7 @@
      "/api/v1/embed/" +
      CONFIG.botUuid +
      "/ws/connect?session_type=" +
-      CONFIG.sessionType +
-      "&session_id=" +
-      encodeURIComponent(state.sessionId);
+      CONFIG.sessionType;

    try {
      state.ws = new WebSocket(url);
@@ -574,8 +520,6 @@
      case "connected":
        state.isConnected = true;
        state.connectionId = data.connection_id;
-        if (state.hasConnected) loadHistory(true);
-        state.hasConnected = true;
        updateStatusDot();
        updateSendBtn();
        break;
@@ -643,7 +587,6 @@

    if (existingIdx >= 0) {
      state.messages[existingIdx] = msg;
-      state.messageVersion++;
      updateMessageEl(existingIdx, msg);
    } else {
      addMessage(msg);
@@ -713,27 +656,16 @@
  }

  // ========== Message History ==========
-  function scheduleHistoryReload() {
-    if (state.historyReloadTimer) clearTimeout(state.historyReloadTimer);
-    state.historyReloadTimer = setTimeout(function () {
-      state.historyReloadTimer = null;
-      loadHistory(true);
-    }, 100);
-  }
-
-  function loadHistory(force) {
-    if (state.historyLoaded && !force) return;
+  function loadHistory() {
+    if (state.historyLoaded) return;
    state.historyLoaded = true;
-    var messageVersion = state.messageVersion;

    var url =
      CONFIG.baseUrl +
      "/api/v1/embed/" +
      CONFIG.botUuid +
      "/messages/" +
-      CONFIG.sessionType +
-      "?session_id=" +
-      encodeURIComponent(state.sessionId);
+      CONFIG.sessionType;
    var headers = {};
    if (state.sessionToken)
      headers["Authorization"] = "Bearer " + state.sessionToken;
@@ -743,16 +675,6 @@
      })
      .then(function (json) {
        if (json.code === 0 && json.data && json.data.messages) {
-          if (force && messageVersion !== state.messageVersion) {
-            scheduleHistoryReload();
-            return;
-          }
-          if (force) {
-            state.messages = [];
-            state.isStreaming = false;
-            state.streamingMsgId = null;
-            renderMessages();
-          }
          var msgs = json.data.messages;
          for (var i = 0; i < msgs.length; i++) {
            addMessage(msgs[i], true);
@@ -771,16 +693,13 @@
      "/api/v1/embed/" +
      CONFIG.botUuid +
      "/reset/" +
-      CONFIG.sessionType +
-      "?session_id=" +
-      encodeURIComponent(state.sessionId);
+      CONFIG.sessionType;
    var headers = {};
    if (state.sessionToken)
      headers["Authorization"] = "Bearer " + state.sessionToken;
    fetch(url, { method: "POST", headers: headers })
      .then(function () {
        state.messages = [];
-        state.messageVersion++;
        state.isStreaming = false;
        state.streamingMsgId = null;
        state.historyLoaded = true;
@@ -794,7 +713,6 @@
  // ========== UI Rendering ==========
  function addMessage(msg, silent) {
    state.messages.push(msg);
-    if (!silent) state.messageVersion++;
    var el = createMessageEl(msg);
    if (els.welcome) {
      els.welcome.style.display = "none";
@@ -15,7 +15,6 @@ from tests.factories import FakeApp


 pytestmark = pytest.mark.integration
-SESSION_ID = '31c0f2e9-b115-4ee6-8f15-3e624d6456b1'


@pytest.fixture(scope='module')
@@ -192,10 +191,10 @@ class TestEmbedMessagesEndpoint:
    """Tests for messages endpoint."""

    @pytest.mark.asyncio
-    async def test_get_messages_person_success(self, quart_test_client, fake_embed_app):
+    async def test_get_messages_person_success(self, quart_test_client):
        """GET messages/person returns messages."""
        response = await quart_test_client.get(
-            f'/api/v1/embed/a1b2c3d4-5678-90ab-cdef-123456789abc/messages/person?session_id={SESSION_ID}',
+            '/api/v1/embed/a1b2c3d4-5678-90ab-cdef-123456789abc/messages/person',
            headers={'Authorization': 'Bearer 1234567890.dummy'},
        )

@@ -203,30 +202,17 @@ class TestEmbedMessagesEndpoint:
        data = await response.get_json()
        assert data['code'] == 0
        assert 'messages' in data['data']
-        fake_embed_app.platform_mgr.websocket_proxy_bot.adapter.get_websocket_messages.assert_called_with(
-            'test-pipeline-uuid', 'person', SESSION_ID
-        )

    @pytest.mark.asyncio
    async def test_get_messages_group_success(self, quart_test_client):
        """GET messages/group returns messages."""
        response = await quart_test_client.get(
-            f'/api/v1/embed/a1b2c3d4-5678-90ab-cdef-123456789abc/messages/group?session_id={SESSION_ID}',
+            '/api/v1/embed/a1b2c3d4-5678-90ab-cdef-123456789abc/messages/group',
            headers={'Authorization': 'Bearer 1234567890.dummy'},
        )

        assert response.status_code == 200

-    @pytest.mark.asyncio
-    async def test_get_messages_requires_session_id(self, quart_test_client):
-        """GET messages without a client session identifier returns 400."""
-        response = await quart_test_client.get(
-            '/api/v1/embed/a1b2c3d4-5678-90ab-cdef-123456789abc/messages/person',
-            headers={'Authorization': 'Bearer 1234567890.dummy'},
-        )
-
-        assert response.status_code == 400
-
    @pytest.mark.asyncio
    async def test_get_messages_invalid_session_type(self, quart_test_client):
        """GET messages with invalid session_type returns 400."""
@@ -243,29 +229,16 @@ class TestEmbedResetEndpoint:
    """Tests for session reset endpoint."""

    @pytest.mark.asyncio
-    async def test_reset_session_person_success(self, quart_test_client, fake_embed_app):
+    async def test_reset_session_person_success(self, quart_test_client):
        """POST reset/person resets session."""
        response = await quart_test_client.post(
-            f'/api/v1/embed/a1b2c3d4-5678-90ab-cdef-123456789abc/reset/person?session_id={SESSION_ID}',
+            '/api/v1/embed/a1b2c3d4-5678-90ab-cdef-123456789abc/reset/person',
            headers={'Authorization': 'Bearer 1234567890.dummy'},
        )

        assert response.status_code == 200
        data = await response.get_json()
        assert data['code'] == 0
-        fake_embed_app.platform_mgr.websocket_proxy_bot.adapter.reset_session.assert_called_with(
-            'test-pipeline-uuid', 'person', SESSION_ID
-        )
-
-    @pytest.mark.asyncio
-    async def test_reset_session_requires_session_id(self, quart_test_client):
-        """POST reset without a client session identifier returns 400."""
-        response = await quart_test_client.post(
-            '/api/v1/embed/a1b2c3d4-5678-90ab-cdef-123456789abc/reset/person',
-            headers={'Authorization': 'Bearer 1234567890.dummy'},
-        )
-
-        assert response.status_code == 400

    @pytest.mark.asyncio
    async def test_reset_session_invalid_uuid(self, quart_test_client):
@@ -77,12 +77,11 @@ async def backend(valkey_config):
        ValkeySearchVectorDatabase,
        VALKEY_SEARCH_AVAILABLE,
    )
+    from glide import ft

    if not VALKEY_SEARCH_AVAILABLE:
        pytest.skip('valkey-glide not installed')

-    from glide import ft
-
    ap = _make_ap(valkey_config)
    db = ValkeySearchVectorDatabase(ap)
    client = await db._ensure_client()
@@ -777,44 +777,6 @@ class TestModelProviderServiceScanProviderModels:
        assert len(result['models']) == 1
        assert result['models'][0]['type'] == 'llm'

-    async def test_scan_provider_marks_existing_rerank_model(self):
-        """Rerank scan results use the rerank service when computing already_added."""
-        ap = SimpleNamespace()
-        ap.persistence_mgr = SimpleNamespace()
-        ap.model_mgr = SimpleNamespace()
-        ap.llm_model_service = SimpleNamespace()
-        ap.embedding_models_service = SimpleNamespace()
-        ap.rerank_models_service = SimpleNamespace()
-
-        provider = _create_mock_provider(provider_uuid='rerank-scan-uuid')
-        ap.persistence_mgr.execute_async = AsyncMock(return_value=_create_mock_result([], first_item=provider))
-        ap.persistence_mgr.serialize_model = Mock(
-            return_value={
-                'uuid': 'rerank-scan-uuid',
-                'name': 'New API',
-                'requester': 'new-api-chat-completions',
-                'base_url': 'https://new-api.example.com/v1',
-                'api_keys': ['key'],
-            }
-        )
-
-        runtime_provider = Mock()
-        runtime_provider.token_mgr.get_token.return_value = 'token'
-        runtime_provider.requester.scan_models = AsyncMock(
-            return_value={'models': [{'id': 'Qwen3-Reranker-8B', 'type': 'rerank'}]}
-        )
-        ap.model_mgr.load_provider = AsyncMock(return_value=runtime_provider)
-        ap.llm_model_service.get_llm_models_by_provider = AsyncMock(return_value=[])
-        ap.embedding_models_service.get_embedding_models_by_provider = AsyncMock(return_value=[])
-        ap.rerank_models_service.get_rerank_models_by_provider = AsyncMock(
-            return_value=[{'name': 'Qwen3-Reranker-8B'}]
-        )
-
-        result = await ModelProviderService(ap).scan_provider_models('rerank-scan-uuid', model_type='rerank')
-
-        assert result['models'][0]['type'] == 'rerank'
-        assert result['models'][0]['already_added'] is True
-
    async def test_scan_provider_not_implemented_raises_error(self):
        """Raises ValueError when scan not implemented."""
        # Setup
@@ -693,7 +693,7 @@ class TestSpaceServiceGetModels:
                            'uuid': 'uuid-2',
                            'model_id': 'model-2',
                            'provider': 'provider-2',
-                            'category': 'rerank',
+                            'category': 'chat',
                            'status': 'active',
                        },
                    ]
@@ -714,7 +714,6 @@ class TestSpaceServiceGetModels:

        # Verify
        assert len(result) == 2
-        assert result[1].category == 'rerank'

    async def test_get_models_api_error(self):
        """Raises ValueError on API error."""
@@ -1,13 +1,11 @@
 from __future__ import annotations

-import asyncio
 from types import SimpleNamespace
-from unittest.mock import AsyncMock, Mock
+from unittest.mock import Mock

 import pytest

 from langbot_plugin.box.client import ActionRPCBoxClient
-from langbot.pkg.box import connector as connector_module
 from langbot.pkg.box.connector import BoxRuntimeConnector


@@ -106,136 +104,3 @@ def test_box_runtime_connector_dispose_terminates_subprocess(monkeypatch: pytest
    ctrl_task.cancel.assert_called_once()
    assert connector._handler_task is None
    assert connector._ctrl_task is None
-
-
-@pytest.mark.asyncio
-async def test_box_runtime_connector_cleans_partial_transport_on_connect_failure(
-    monkeypatch: pytest.MonkeyPatch,
-):
-    monkeypatch.setattr('langbot.pkg.utils.platform.get_platform', lambda: 'linux')
-    monkeypatch.setattr('langbot.pkg.utils.platform.standalone_box', False)
-    connector = BoxRuntimeConnector(make_app(Mock()))
-    connector._start_local_stdio = AsyncMock(side_effect=RuntimeError('bind failed'))
-    connector._stop_transport = AsyncMock()
-    connector._close_managed_subprocess = AsyncMock()
-
-    with pytest.raises(RuntimeError, match='bind failed'):
-        await connector.initialize()
-
-    assert connector._stop_transport.await_count == 2
-    connector._close_managed_subprocess.assert_awaited_once()
-
-
-@pytest.mark.asyncio
-async def test_box_runtime_connector_starts_heartbeat_after_reconnect(
-    monkeypatch: pytest.MonkeyPatch,
-):
-    monkeypatch.setattr('langbot.pkg.utils.platform.get_platform', lambda: 'linux')
-    monkeypatch.setattr('langbot.pkg.utils.platform.standalone_box', False)
-    connector = BoxRuntimeConnector(make_app(Mock()))
-    connector._start_local_stdio = AsyncMock(side_effect=[RuntimeError('bind failed'), None])
-    connector._stop_transport = AsyncMock()
-    connector._close_managed_subprocess = AsyncMock()
-
-    with pytest.raises(RuntimeError, match='bind failed'):
-        await connector.initialize()
-
-    assert connector._heartbeat_task is None
-
-    await connector.reconnect()
-
-    assert connector._heartbeat_task is not None
-    assert not connector._heartbeat_task.done()
-    await connector.aclose()
-
-
-@pytest.mark.asyncio
-async def test_box_stdio_connection_does_not_capture_unconsumed_stderr(
-    monkeypatch: pytest.MonkeyPatch,
-):
-    monkeypatch.setattr('langbot.pkg.utils.platform.get_platform', lambda: 'linux')
-    monkeypatch.setattr('langbot.pkg.utils.platform.standalone_box', False)
-    created = {}
-
-    class FakeHandler:
-        def __init__(self, connection):
-            self.release = asyncio.Event()
-
-        async def call_action(self, action, data):
-            return None
-
-        async def run(self):
-            await self.release.wait()
-
-        async def close(self):
-            self.release.set()
-
-    class FakeController:
-        def __init__(self, **kwargs):
-            created.update(kwargs)
-            self.process = SimpleNamespace(returncode=0)
-
-        async def run(self, callback):
-            await callback(object())
-
-        async def close(self):
-            return None
-
-    monkeypatch.setattr(connector_module, 'Handler', FakeHandler)
-    monkeypatch.setattr(
-        'langbot_plugin.runtime.io.controllers.stdio.client.StdioClientController',
-        FakeController,
-    )
-    connector = BoxRuntimeConnector(make_app(Mock()))
-
-    await connector.initialize()
-
-    assert created['capture_stderr'] is False
-    assert connector._handler is not None
-    await connector.aclose()
-
-
-@pytest.mark.asyncio
-async def test_box_disconnect_notifies_once_and_clears_handler(
-    monkeypatch: pytest.MonkeyPatch,
-):
-    monkeypatch.setattr('langbot.pkg.utils.platform.get_platform', lambda: 'linux')
-    monkeypatch.setattr('langbot.pkg.utils.platform.standalone_box', False)
-    disconnect = AsyncMock()
-
-    class FakeHandler:
-        def __init__(self, connection):
-            pass
-
-        async def call_action(self, action, data):
-            return None
-
-        async def run(self):
-            return None
-
-        async def close(self):
-            return None
-
-    class FakeController:
-        def __init__(self, **kwargs):
-            self.process = SimpleNamespace(returncode=0)
-
-        async def run(self, callback):
-            await callback(object())
-
-        async def close(self):
-            return None
-
-    monkeypatch.setattr(connector_module, 'Handler', FakeHandler)
-    monkeypatch.setattr(
-        'langbot_plugin.runtime.io.controllers.stdio.client.StdioClientController',
-        FakeController,
-    )
-    connector = BoxRuntimeConnector(make_app(Mock()), runtime_disconnect_callback=disconnect)
-
-    await connector.initialize()
-    await asyncio.sleep(0)
-
-    disconnect.assert_awaited_once_with(connector)
-    assert connector._handler is None
-    await connector.aclose()
@@ -325,50 +325,10 @@ async def test_box_service_dispose_schedules_shutdown_on_event_loop(monkeypatch:
    service.dispose()
    await asyncio.sleep(0)

-    connector.dispose.assert_not_called()
+    connector.dispose.assert_called_once()
    service.shutdown.assert_awaited_once()


-@pytest.mark.asyncio
-async def test_box_service_shutdown_reaps_connector_when_runtime_rpc_is_offline(
-    monkeypatch: pytest.MonkeyPatch,
-):
-    connector = Mock()
-    connector.client = Mock()
-    connector.client.shutdown = AsyncMock(side_effect=RuntimeError('offline'))
-    connector.aclose = AsyncMock()
-
-    monkeypatch.setattr('langbot.pkg.box.service.BoxRuntimeConnector', Mock(return_value=connector))
-
-    service = BoxService(make_app(Mock()))
-    await service.shutdown()
-
-    connector.client.shutdown.assert_awaited_once()
-    connector.aclose.assert_awaited_once()
-
-
-@pytest.mark.asyncio
-async def test_box_service_reconnect_restores_workspace_and_runs_cleanup(
-    monkeypatch: pytest.MonkeyPatch,
-):
-    app = make_app(Mock())
-    app.skill_mgr = SimpleNamespace(reload_skills=AsyncMock())
-    service = BoxService(app, client=Mock(spec=BoxRuntimeClient))
-    connector = Mock()
-    connector.reconnect = AsyncMock()
-    service._ensure_default_workspace = Mock()
-    service._purge_attachment_dirs = AsyncMock()
-    monkeypatch.setattr('langbot.pkg.box.service.asyncio.sleep', AsyncMock())
-
-    await service._reconnect_loop(connector)
-
-    connector.reconnect.assert_awaited_once()
-    service._ensure_default_workspace.assert_called_once()
-    service._purge_attachment_dirs.assert_awaited_once()
-    app.skill_mgr.reload_skills.assert_awaited_once()
-    assert service.available is True
-
-
@pytest.mark.asyncio
 async def test_box_runtime_reuses_request_session():
    logger = Mock()
@@ -2,7 +2,6 @@ from __future__ import annotations

 import signal
 from types import SimpleNamespace
-from unittest.mock import Mock

 import pytest

@@ -19,60 +18,47 @@ async def test_main_signal_handler_handles_sigint_before_app_created(monkeypatch
    async def fake_make_app(loop):
        captured_handler[signal.SIGINT](signal.SIGINT, None)

+    def fake_exit(code):
+        raise SystemExit(code)
+
    monkeypatch.setattr(signal, 'signal', fake_signal)
    monkeypatch.setattr(boot, 'make_app', fake_make_app)
+    monkeypatch.setattr(boot.os, '_exit', fake_exit)

-    await boot.main(SimpleNamespace())
+    with pytest.raises(SystemExit) as exc_info:
+        await boot.main(SimpleNamespace())
+
+    assert exc_info.value.code == 0


@pytest.mark.asyncio
 async def test_main_signal_handler_disposes_created_app(monkeypatch):
    captured_handler = {}
-    app_inst = SimpleNamespace(shutdown_called=False)
+    app_inst = SimpleNamespace(disposed=False)

    def fake_signal(sig, handler):
        captured_handler[sig] = handler

-    async def shutdown():
-        app_inst.shutdown_called = True
+    def dispose():
+        app_inst.disposed = True

    async def run():
        captured_handler[signal.SIGINT](signal.SIGINT, None)

    async def fake_make_app(loop):
-        app_inst.shutdown = shutdown
+        app_inst.dispose = dispose
        app_inst.run = run
        return app_inst

+    def fake_exit(code):
+        raise SystemExit(code)
+
    monkeypatch.setattr(signal, 'signal', fake_signal)
    monkeypatch.setattr(boot, 'make_app', fake_make_app)
+    monkeypatch.setattr(boot.os, '_exit', fake_exit)

-    await boot.main(SimpleNamespace())
+    with pytest.raises(SystemExit) as exc_info:
+        await boot.main(SimpleNamespace())

-    assert app_inst.shutdown_called is True
-
-
-@pytest.mark.asyncio
-async def test_main_reports_app_run_failure_and_still_shuts_down(monkeypatch):
-    app_inst = SimpleNamespace(shutdown_called=False)
-
-    async def shutdown():
-        app_inst.shutdown_called = True
-
-    async def run():
-        raise RuntimeError('run failed')
-
-    async def fake_make_app(loop):
-        app_inst.shutdown = shutdown
-        app_inst.run = run
-        return app_inst
-
-    print_exc = Mock()
-    monkeypatch.setattr(signal, 'signal', lambda *_args: None)
-    monkeypatch.setattr(boot, 'make_app', fake_make_app)
-    monkeypatch.setattr(boot.traceback, 'print_exc', print_exc)
-
-    await boot.main(SimpleNamespace())
-
-    print_exc.assert_called_once()
-    assert app_inst.shutdown_called is True
+    assert exc_info.value.code == 0
+    assert app_inst.disposed is True
@@ -158,21 +158,17 @@ class TestCommandHandlerReal:

    @pytest.mark.asyncio
    async def test_admin_privilege_check(self, fake_app, mock_event_ctx, mock_execute_factory):
-        """A per-bot admin from the database is marked as admin in command events."""
+        """Admin users get privilege level 2."""
        from langbot_plugin.api.entities.builtin.provider.session import LauncherTypes

        command = get_command_handler()

-        admin_result = Mock()
-        admin_result.first.return_value = Mock()
-        fake_app.persistence_mgr.execute_async = AsyncMock(return_value=admin_result)
-        fake_app.instance_config.data = {}
+        fake_app.instance_config.data = {'admins': ['person_12345']}
        fake_app.plugin_connector.emit_event = AsyncMock(return_value=mock_event_ctx)
        fake_app.cmd_mgr.execute = mock_execute_factory()

        handler = command.CommandHandler(fake_app)
        query = command_query('status')
-        query.bot_uuid = 'bot-1'
        query.launcher_type = LauncherTypes.PERSON
        query.launcher_id = 12345

@@ -180,28 +176,23 @@ class TestCommandHandlerReal:
        async for result in handler.handle(query):
            results.append(result)

-        fake_app.persistence_mgr.execute_async.assert_awaited_once()
        call_args = fake_app.plugin_connector.emit_event.call_args
        event = call_args[0][0]
        assert event.is_admin is True

    @pytest.mark.asyncio
    async def test_non_admin_privilege_check(self, fake_app, mock_event_ctx, mock_execute_factory):
-        """A launcher absent from the per-bot admin table is not an admin."""
+        """Non-admin users get privilege level 1."""
        from langbot_plugin.api.entities.builtin.provider.session import LauncherTypes

        command = get_command_handler()

-        admin_result = Mock()
-        admin_result.first.return_value = None
-        fake_app.persistence_mgr.execute_async = AsyncMock(return_value=admin_result)
-        fake_app.instance_config.data = {}
+        fake_app.instance_config.data = {'admins': ['person_12345']}
        fake_app.plugin_connector.emit_event = AsyncMock(return_value=mock_event_ctx)
        fake_app.cmd_mgr.execute = mock_execute_factory()

        handler = command.CommandHandler(fake_app)
        query = command_query('status')
-        query.bot_uuid = 'bot-1'
        query.launcher_type = LauncherTypes.PERSON
        query.launcher_id = 67890

@@ -209,7 +200,6 @@ class TestCommandHandlerReal:
        async for result in handler.handle(query):
            results.append(result)

-        fake_app.persistence_mgr.execute_async.assert_awaited_once()
        call_args = fake_app.plugin_connector.emit_event.call_args
        event = call_args[0][0]
        assert event.is_admin is False
@@ -2,7 +2,6 @@ import pytest
 import aiocqhttp

 import langbot_plugin.api.entities.builtin.platform.message as platform_message
-import langbot_plugin.api.entities.builtin.platform.events as platform_events
 from langbot.pkg.platform.sources.aiocqhttp import (
    AiocqhttpAdapter,
    AiocqhttpEventConverter,
@@ -16,72 +15,6 @@ async def _convert_single(component: platform_message.MessageComponent):
    return message[0]


-class _TestLogger:
-    def __init__(self):
-        self.messages = []
-
-    async def info(self, message):
-        self.messages.append(message)
-
-
-def _make_adapter():
-    logger = _TestLogger()
-    adapter = AiocqhttpAdapter.model_construct(
-        config={},
-        logger=logger,
-        bot=aiocqhttp.CQHttp(),
-        on_websocket_connection_event_cache=[],
-        _listener_wrappers={},
-    )
-    adapter.bot.on_websocket_connection(adapter._on_websocket_connection)
-    return adapter, logger
-
-
-def test_connection_listener_is_registered_once_for_multiple_message_listeners():
-    adapter, _ = _make_adapter()
-
-    async def callback(event, source_adapter):
-        return None
-
-    adapter.register_listener(platform_events.FriendMessage, callback)
-    adapter.register_listener(platform_events.GroupMessage, callback)
-    adapter.register_listener(platform_events.FeedbackEvent, callback)
-
-    assert len(adapter.bot._bus._subscribers['meta_event.lifecycle.connect']) == 1
-
-
-@pytest.mark.asyncio
-async def test_connection_listener_only_suppresses_exact_duplicates():
-    adapter, logger = _make_adapter()
-    first = aiocqhttp.Event({'self_id': 1001, 'time': 10})
-    duplicate = aiocqhttp.Event({'self_id': 1001, 'time': 10})
-    second = aiocqhttp.Event({'self_id': 2002, 'time': 20})
-
-    await adapter._on_websocket_connection(first)
-    await adapter._on_websocket_connection(duplicate)
-    await adapter._on_websocket_connection(second)
-
-    assert adapter.on_websocket_connection_event_cache == [first, second]
-    assert logger.messages == [
-        'WebSocket connection established, bot id: 1001',
-        'WebSocket connection established, bot id: 2002',
-    ]
-
-
-def test_unregister_listener_removes_registered_wrapper():
-    adapter, _ = _make_adapter()
-
-    async def callback(event, source_adapter):
-        return None
-
-    adapter.register_listener(platform_events.GroupMessage, callback)
-    assert len(adapter.bot._bus._subscribers['message.group']) == 1
-
-    adapter.unregister_listener(platform_events.GroupMessage, callback)
-
-    assert not adapter.bot._bus._subscribers['message.group']
-
-
@pytest.mark.asyncio
@pytest.mark.parametrize(
    ('payload', 'expected'),
@@ -1,39 +0,0 @@
-import asyncio
-import threading
-
-import pytest
-
-from langbot.pkg.platform.sources.lark import NonBlockingLarkWSClient
-
-
-@pytest.mark.asyncio
-async def test_lark_connection_url_lookup_does_not_block_main_event_loop(monkeypatch):
-    client = NonBlockingLarkWSClient('app-id', 'app-secret')
-    lookup_started = threading.Event()
-    release_lookup = threading.Event()
-    base_connect_urls: list[str] = []
-
-    def blocking_get_conn_url() -> str:
-        lookup_started.set()
-        if not release_lookup.wait(timeout=2):
-            raise TimeoutError('test did not release the connection lookup')
-        return 'wss://example.invalid/connect?device_id=device&service_id=service'
-
-    async def fake_base_connect(self):
-        base_connect_urls.append(self._get_conn_url())
-
-    monkeypatch.setattr(client, '_get_conn_url', blocking_get_conn_url)
-    monkeypatch.setattr('lark_oapi.ws.Client._connect', fake_base_connect)
-
-    connect_task = asyncio.create_task(client._connect())
-    await asyncio.wait_for(asyncio.to_thread(lookup_started.wait, 1), timeout=1)
-
-    # If the SDK's synchronous requests.post still runs on the event-loop
-    # thread, this sleep cannot complete until release_lookup is set.
-    await asyncio.wait_for(asyncio.sleep(0.01), timeout=0.1)
-    assert not connect_task.done()
-
-    release_lookup.set()
-    await asyncio.wait_for(connect_task, timeout=1)
-
-    assert base_connect_urls == ['wss://example.invalid/connect?device_id=device&service_id=service']
@@ -1,267 +0,0 @@
-"""Regression tests for isolated embed-widget conversations."""
-
-import asyncio
-from pathlib import Path
-from unittest.mock import AsyncMock, Mock
-
-import pytest
-
-import langbot_plugin.api.entities.builtin.platform.events as platform_events
-from langbot.pkg.platform.sources import websocket_adapter as websocket_adapter_module
-from langbot.pkg.platform.sources.websocket_adapter import WebSocketAdapter, WebSocketMessage, WebSocketSession
-from langbot.pkg.platform.sources.websocket_manager import WebSocketConnectionManager, is_valid_session_id
-
-
-@pytest.mark.asyncio
-async def test_broadcast_only_reaches_connections_in_same_browser_session():
-    manager = WebSocketConnectionManager()
-    first = await manager.add_connection(
-        websocket=Mock(),
-        pipeline_uuid='pipeline-1',
-        session_type='person',
-        session_id='session-a',
-    )
-    second = await manager.add_connection(
-        websocket=Mock(),
-        pipeline_uuid='pipeline-1',
-        session_type='person',
-        session_id='session-b',
-    )
-    dashboard = await manager.add_connection(
-        websocket=Mock(),
-        pipeline_uuid='pipeline-1',
-        session_type='person',
-    )
-
-    await manager.broadcast_to_pipeline(
-        'pipeline-1',
-        {'type': 'response'},
-        session_type='person',
-        session_id='session-a',
-    )
-
-    assert await first.send_queue.get() == {'type': 'response'}
-    assert second.send_queue.empty()
-    assert dashboard.send_queue.empty()
-
-    await manager.broadcast_to_pipeline(
-        'pipeline-1',
-        {'type': 'dashboard-response'},
-        session_type='person',
-        session_id=None,
-    )
-
-    assert await dashboard.send_queue.get() == {'type': 'dashboard-response'}
-    assert first.send_queue.empty()
-    assert second.send_queue.empty()
-
-
-@pytest.mark.asyncio
-async def test_embed_event_uses_stable_session_launcher(monkeypatch):
-    manager = WebSocketConnectionManager()
-    session_id = '31c0f2e9-b115-4ee6-8f15-3e624d6456b1'
-    connection = await manager.add_connection(
-        websocket=Mock(),
-        pipeline_uuid='pipeline-1',
-        session_type='person',
-        session_id=session_id,
-    )
-    monkeypatch.setattr(websocket_adapter_module, 'ws_connection_manager', manager)
-
-    adapter = WebSocketAdapter.model_construct(ap=Mock(), logger=AsyncMock())
-    adapter.websocket_person_session = WebSocketSession(id='person')
-    adapter.websocket_group_session = WebSocketSession(id='group')
-    received = []
-
-    async def listener(event, _callback_adapter):
-        received.append(event)
-
-    adapter.listeners = {platform_events.FriendMessage: listener}
-    await adapter.handle_websocket_message(
-        connection,
-        {'message': [{'type': 'Plain', 'text': 'hello'}], 'stream': False},
-    )
-    await asyncio.sleep(0)
-
-    assert received[0].sender.id == f'websocket_pipeline-1:{session_id}'
-
-
-@pytest.mark.asyncio
-async def test_embed_group_event_uses_stable_session_launcher(monkeypatch):
-    manager = WebSocketConnectionManager()
-    session_id = '31c0f2e9-b115-4ee6-8f15-3e624d6456b1'
-    connection = await manager.add_connection(
-        websocket=Mock(),
-        pipeline_uuid='pipeline-1',
-        session_type='group',
-        session_id=session_id,
-    )
-    monkeypatch.setattr(websocket_adapter_module, 'ws_connection_manager', manager)
-
-    adapter = WebSocketAdapter.model_construct(ap=Mock(), logger=AsyncMock())
-    adapter.websocket_person_session = WebSocketSession(id='person')
-    adapter.websocket_group_session = WebSocketSession(id='group')
-    received = []
-
-    async def listener(event, _callback_adapter):
-        received.append(event)
-
-    adapter.listeners = {platform_events.GroupMessage: listener}
-    await adapter.handle_websocket_message(
-        connection,
-        {'message': [{'type': 'Plain', 'text': 'hello'}], 'stream': False},
-    )
-    await asyncio.sleep(0)
-
-    assert received[0].sender.id == f'websocket_pipeline-1:{session_id}'
-    assert received[0].sender.group.id == f'websocketgroup_pipeline-1:{session_id}'
-
-    dashboard = await manager.add_connection(
-        websocket=Mock(),
-        pipeline_uuid='pipeline-1',
-        session_type='group',
-    )
-    await adapter.handle_websocket_message(
-        dashboard,
-        {'stream': False, 'message': [{'type': 'Plain', 'text': 'dashboard'}]},
-        None,
-    )
-    await asyncio.sleep(0)
-
-    assert received[1].sender.id == f'websocket_{dashboard.connection_id}'
-    assert received[1].sender.group.id == 'websocketgroup'
-
-
-@pytest.mark.asyncio
-async def test_stable_session_launcher_resolves_to_active_connection(monkeypatch):
-    manager = WebSocketConnectionManager()
-    session_id = '31c0f2e9-b115-4ee6-8f15-3e624d6456b1'
-    await manager.add_connection(
-        websocket=Mock(),
-        pipeline_uuid='pipeline-2',
-        session_type='person',
-        session_id=session_id,
-    )
-    connection = await manager.add_connection(
-        websocket=Mock(),
-        pipeline_uuid='pipeline-1',
-        session_type='person',
-        session_id=session_id,
-    )
-    monkeypatch.setattr(websocket_adapter_module, 'ws_connection_manager', manager)
-
-    adapter = WebSocketAdapter.model_construct(ap=Mock(), logger=AsyncMock())
-    message_source = Mock()
-    message_source.sender.id = f'websocket_pipeline-1:{session_id}'
-
-    assert await adapter._get_message_context(message_source) == ('pipeline-1', session_id)
-    assert await adapter._get_connection_from_target(f'websocketgroup_pipeline-1:{session_id}') is connection
-    assert await manager.get_connection_by_session_id(session_id, 'pipeline-1') is connection
-
-    await manager.remove_connection(connection.connection_id)
-
-    assert await adapter._get_message_context(message_source) == ('pipeline-1', session_id)
-    assert await manager.get_connection_by_session_id(session_id, 'pipeline-1') is None
-
-
-def test_session_ids_must_be_canonical_random_uuids():
-    assert is_valid_session_id('31c0f2e9-b115-4ee6-8f15-3e624d6456b1')
-    assert not is_valid_session_id('session-a')
-    assert not is_valid_session_id('00000000-0000-0000-0000-000000000000')
-
-
-def test_history_read_does_not_allocate_unknown_session():
-    adapter = WebSocketAdapter.model_construct(ap=Mock(), logger=AsyncMock())
-    adapter.websocket_person_session = WebSocketSession(id='person')
-    adapter.websocket_group_session = WebSocketSession(id='group')
-
-    assert adapter.get_websocket_messages('pipeline-1', 'person', 'missing-session') == []
-    assert adapter.websocket_person_session.message_lists == {}
-
-
-def test_history_and_reset_are_scoped_to_browser_session():
-    matching_provider_session = Mock(
-        launcher_type=Mock(value='person'),
-        launcher_id='websocket_pipeline-1:session-a',
-    )
-    matching_group_provider_session = Mock(
-        launcher_type=Mock(value='group'),
-        launcher_id='websocketgroup_pipeline-1:session-a',
-    )
-    other_session = Mock(
-        launcher_type=Mock(value='person'),
-        launcher_id='websocket_pipeline-1:session-b',
-    )
-    ap = Mock()
-    ap.sess_mgr.session_list = [
-        matching_provider_session,
-        matching_group_provider_session,
-        other_session,
-    ]
-    adapter = WebSocketAdapter.model_construct(
-        ap=ap,
-        logger=AsyncMock(),
-    )
-    adapter.websocket_person_session = Mock()
-    adapter.websocket_group_session = Mock()
-
-    session_a = [
-        WebSocketMessage(
-            id=1,
-            role='user',
-            content='private-a',
-            message_chain=[],
-            timestamp='2026-07-13T00:00:00',
-        )
-    ]
-    session_b = [
-        WebSocketMessage(
-            id=1,
-            role='user',
-            content='private-b',
-            message_chain=[],
-            timestamp='2026-07-13T00:00:00',
-        )
-    ]
-    histories = {
-        'pipeline-1:session-a': session_a,
-        'pipeline-1:session-b': session_b,
-    }
-    stream_indexes = {
-        'pipeline-1:session-a': {'response-a': 0},
-        'pipeline-1:session-b': {'response-b': 0},
-    }
-    adapter.websocket_person_session.get_message_list.side_effect = histories.__getitem__
-    adapter.websocket_person_session.message_lists = histories
-    adapter.websocket_person_session.stream_message_indexes = stream_indexes
-    adapter.websocket_group_session.message_lists = {}
-    adapter.websocket_group_session.stream_message_indexes = {}
-
-    assert adapter.get_websocket_messages('pipeline-1', 'person', 'session-a')[0]['content'] == 'private-a'
-    assert adapter.get_websocket_messages('pipeline-1', 'person', 'session-b')[0]['content'] == 'private-b'
-
-    adapter.reset_session('pipeline-1', 'person', 'session-a')
-
-    assert histories['pipeline-1:session-a'] == []
-    assert stream_indexes['pipeline-1:session-a'] == {}
-    assert histories['pipeline-1:session-b'] == session_b
-    assert stream_indexes['pipeline-1:session-b'] == {'response-b': 0}
-    assert ap.sess_mgr.session_list == [matching_group_provider_session, other_session]
-
-    adapter.reset_session('pipeline-1', 'group', 'session-a')
-
-    assert ap.sess_mgr.session_list == [other_session]
-
-
-def test_widget_sends_stable_session_id_to_all_conversation_endpoints():
-    widget_path = Path(__file__).parents[3] / 'src/langbot/templates/embed/widget.js'
-    widget = widget_path.read_text(encoding='utf-8')
-
-    assert 'langbot_embed_session_' in widget
-    assert 'window.sessionStorage' in widget
-    assert 'window.localStorage' not in widget
-    assert 'session_id=' in widget
-    assert widget.count('encodeURIComponent(state.sessionId)') >= 3
-    assert 'loadHistory(true)' in widget
-    assert 'messageVersion !== state.messageVersion' in widget
-    assert 'scheduleHistoryReload();' in widget
@@ -62,25 +62,6 @@ class TestListPlugins:

        assert result == []

-    @pytest.mark.asyncio
-    async def test_returns_empty_while_runtime_is_disconnected(self):
-        connector = create_mock_connector()
-
-        result = await connector.list_plugins()
-
-        assert result == []
-
-    @pytest.mark.asyncio
-    async def test_returns_empty_after_managed_transport_disconnects(self):
-        connector = create_mock_connector()
-        connector.handler = AsyncMock()
-        connector._transport_task = Mock()
-
-        result = await connector.list_plugins()
-
-        assert result == []
-        connector.handler.list_plugins.assert_not_called()
-
    @pytest.mark.asyncio
    async def test_calls_handler_list_plugins(self):
        """Test that handler.list_plugins is called."""
@@ -313,12 +294,6 @@ class TestListKnowledgeEngines:
        connector.handler.list_knowledge_engines.assert_called_once()
        assert result == [{'plugin_id': 'author/engine', 'name': 'Engine'}]

-    @pytest.mark.asyncio
-    async def test_returns_empty_while_runtime_is_disconnected(self):
-        connector = create_mock_connector()
-
-        assert await connector.list_knowledge_engines() == []
-

 class TestListParsers:
    """Tests for list_parsers method."""
@@ -356,12 +331,6 @@ class TestListParsers:
        connector.handler.list_parsers.assert_called_once()
        assert result == [{'plugin_id': 'author/parser', 'supported_mime_types': ['text/plain']}]

-    @pytest.mark.asyncio
-    async def test_returns_empty_while_runtime_is_disconnected(self):
-        connector = create_mock_connector()
-
-        assert await connector.list_parsers() == []
-

 class TestCallParser:
    """Tests for call_parser method."""
@@ -1,20 +1,15 @@
 from __future__ import annotations

-import asyncio
 from types import SimpleNamespace
-from unittest.mock import AsyncMock, Mock
+from unittest.mock import AsyncMock

 import pytest

-from langbot.pkg.plugin import connector as connector_module
 from langbot.pkg.plugin.connector import PluginRuntimeConnector, PluginRuntimeNotConnectedError


 def make_connector() -> PluginRuntimeConnector:
-    app = SimpleNamespace(
-        logger=Mock(),
-        instance_config=SimpleNamespace(data={'plugin': {'enable': True}, 'space': {'url': ''}}),
-    )
+    app = SimpleNamespace(instance_config=SimpleNamespace(data={'plugin': {'enable': True}}))
    return PluginRuntimeConnector(app, AsyncMock())


@@ -35,133 +30,3 @@ async def test_ping_plugin_runtime_delegates_to_connected_handler():

    assert result == 'pong'
    connector.handler.ping.assert_awaited_once()
-
-
-@pytest.mark.asyncio
-async def test_stop_transport_tolerates_handler_callback_removing_attribute():
-    connector = make_connector()
-
-    class Handler:
-        async def close(self):
-            del connector.handler
-
-    connector.handler = Handler()
-
-    await connector._stop_transport()
-
-    assert not hasattr(connector, 'handler')
-
-
-@pytest.mark.asyncio
-async def test_stdio_runtime_connection_does_not_capture_unconsumed_stderr(
-    monkeypatch: pytest.MonkeyPatch,
-):
-    connector = make_connector()
-    created = {}
-
-    class FakeRuntimeHandler:
-        def __init__(self, connection, disconnect_callback, ap):
-            self.release = asyncio.Event()
-
-        async def ping(self):
-            return None
-
-        async def set_runtime_config(self, **kwargs):
-            return None
-
-        async def run(self):
-            await self.release.wait()
-
-        async def close(self):
-            self.release.set()
-
-    class FakeController:
-        def __init__(self, **kwargs):
-            created.update(kwargs)
-
-        async def run(self, callback):
-            await callback(object())
-
-        async def close(self):
-            return None
-
-    monkeypatch.setattr(connector_module.platform, 'get_platform', lambda: 'linux')
-    monkeypatch.setattr(
-        connector_module.platform,
-        'use_websocket_to_connect_plugin_runtime',
-        lambda: False,
-    )
-    monkeypatch.setattr(
-        connector_module.handler,
-        'RuntimeConnectionHandler',
-        FakeRuntimeHandler,
-    )
-    monkeypatch.setattr(
-        connector_module.stdio_client_controller,
-        'StdioClientController',
-        FakeController,
-    )
-
-    await connector.initialize()
-
-    assert created['capture_stderr'] is False
-    assert connector._connected.is_set()
-    await connector.aclose()
-
-
-@pytest.mark.asyncio
-async def test_runtime_disconnect_notifies_once_and_clears_handler(
-    monkeypatch: pytest.MonkeyPatch,
-):
-    disconnect = AsyncMock()
-    connector = PluginRuntimeConnector(make_connector().ap, disconnect)
-
-    class FakeRuntimeHandler:
-        def __init__(self, connection, disconnect_callback, ap):
-            self.disconnect_callback = disconnect_callback
-
-        async def ping(self):
-            return None
-
-        async def set_runtime_config(self, **kwargs):
-            return None
-
-        async def run(self):
-            await self.disconnect_callback(self)
-
-        async def close(self):
-            return None
-
-    class FakeController:
-        def __init__(self, **kwargs):
-            pass
-
-        async def run(self, callback):
-            await callback(object())
-
-        async def close(self):
-            return None
-
-    monkeypatch.setattr(connector_module.platform, 'get_platform', lambda: 'linux')
-    monkeypatch.setattr(
-        connector_module.platform,
-        'use_websocket_to_connect_plugin_runtime',
-        lambda: False,
-    )
-    monkeypatch.setattr(
-        connector_module.handler,
-        'RuntimeConnectionHandler',
-        FakeRuntimeHandler,
-    )
-    monkeypatch.setattr(
-        connector_module.stdio_client_controller,
-        'StdioClientController',
-        FakeController,
-    )
-
-    await connector.initialize()
-    await asyncio.sleep(0)
-
-    disconnect.assert_awaited_once_with(connector)
-    assert not hasattr(connector, 'handler')
-    await connector.aclose()
@@ -279,122 +279,6 @@ class TestInvokeLLMStreamUsage:

        assert query.variables['_stream_usage']['total_tokens'] == 12

-    @pytest.mark.asyncio
-    async def test_stream_removes_leading_think_across_chunks(self):
-        """A leading think block split across chunks must be removed."""
-        import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
-        import langbot_plugin.api.entities.builtin.provider.message as provider_message
-
-        mock_ap = Mock()
-        mock_ap.tool_mgr = Mock()
-        mock_ap.tool_mgr.generate_tools_for_openai = AsyncMock(return_value=None)
-        requester = litellmchat.LiteLLMRequester(ap=mock_ap, config={})
-        model = MockRuntimeModel('minimax-m3', 'test-api-key')
-
-        chunks = [
-            self._make_chunk(content='<thi'),
-            self._make_chunk(content='nk>hidden'),
-            self._make_chunk(content=' reasoning</thi'),
-            self._make_chunk(content='nk>Visible answer', finish_reason='stop'),
-        ]
-
-        async def _aiter(*args, **kwargs):
-            for c in chunks:
-                yield c
-
-        query = Mock(spec=pipeline_query.Query)
-        query.variables = {}
-        messages = [provider_message.Message(role='user', content='Hi')]
-
-        with patch.object(litellmchat, 'acompletion', new=AsyncMock(side_effect=lambda **kw: _aiter())):
-            collected = [
-                chunk
-                async for chunk in requester.invoke_llm_stream(
-                    query=query,
-                    model=model,
-                    messages=messages,
-                    remove_think=True,
-                )
-            ]
-
-        assert ''.join(chunk.content or '' for chunk in collected) == 'Visible answer'
-
-    @pytest.mark.asyncio
-    async def test_stream_removes_initial_orphan_think_close(self):
-        """Initial reasoning content without an open tag is removed until </think>."""
-        import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
-        import langbot_plugin.api.entities.builtin.provider.message as provider_message
-
-        mock_ap = Mock()
-        mock_ap.tool_mgr = Mock()
-        mock_ap.tool_mgr.generate_tools_for_openai = AsyncMock(return_value=None)
-        requester = litellmchat.LiteLLMRequester(ap=mock_ap, config={})
-        model = MockRuntimeModel('minimax-m3', 'test-api-key')
-
-        chunks = [
-            self._make_chunk(content='hidden reasoning'),
-            self._make_chunk(content=' still hidden</thi'),
-            self._make_chunk(content='nk>Visible answer', finish_reason='stop'),
-        ]
-
-        async def _aiter(*args, **kwargs):
-            for c in chunks:
-                yield c
-
-        query = Mock(spec=pipeline_query.Query)
-        query.variables = {}
-        messages = [provider_message.Message(role='user', content='Hi')]
-
-        with patch.object(litellmchat, 'acompletion', new=AsyncMock(side_effect=lambda **kw: _aiter())):
-            collected = [
-                chunk
-                async for chunk in requester.invoke_llm_stream(
-                    query=query,
-                    model=model,
-                    messages=messages,
-                    remove_think=True,
-                )
-            ]
-
-        assert ''.join(chunk.content or '' for chunk in collected) == 'Visible answer'
-
-    @pytest.mark.asyncio
-    async def test_stream_removes_non_leading_think_content(self):
-        """A think block in the answer body is removed with its content."""
-        import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
-        import langbot_plugin.api.entities.builtin.provider.message as provider_message
-
-        mock_ap = Mock()
-        mock_ap.tool_mgr = Mock()
-        mock_ap.tool_mgr.generate_tools_for_openai = AsyncMock(return_value=None)
-        requester = litellmchat.LiteLLMRequester(ap=mock_ap, config={})
-        model = MockRuntimeModel('gpt-4o', 'test-api-key')
-
-        chunks = [
-            self._make_chunk(content='Use <think>x</think> as an XML-like example.', finish_reason='stop'),
-        ]
-
-        async def _aiter(*args, **kwargs):
-            for c in chunks:
-                yield c
-
-        query = Mock(spec=pipeline_query.Query)
-        query.variables = {}
-        messages = [provider_message.Message(role='user', content='Hi')]
-
-        with patch.object(litellmchat, 'acompletion', new=AsyncMock(side_effect=lambda **kw: _aiter())):
-            collected = [
-                chunk
-                async for chunk in requester.invoke_llm_stream(
-                    query=query,
-                    model=model,
-                    messages=messages,
-                    remove_think=True,
-                )
-            ]
-
-        assert ''.join(chunk.content or '' for chunk in collected) == 'Use  as an XML-like example.'
-
    @pytest.mark.asyncio
    async def test_stream_tool_call_delta_missing_id_and_name(self):
        """LiteLLM may stream tool-call argument deltas with id/name set to None."""
@@ -598,38 +482,6 @@ class TestProcessThinkingContent:
        result = requester._process_thinking_content(content, None, remove_think=True)
        assert result == 'The answer is 42.'

-    def test_remove_leading_think_tag(self):
-        """Test removing a leading <think> block when remove_think=True"""
-        requester = litellmchat.LiteLLMRequester(ap=Mock(), config={})
-
-        content = '<think>Let me think...</think> The answer is 42.'
-        result = requester._process_thinking_content(content, None, remove_think=True)
-        assert result == 'The answer is 42.'
-
-    def test_remove_non_leading_think_tag(self):
-        """Test removing <think> and its content in the answer body"""
-        requester = litellmchat.LiteLLMRequester(ap=Mock(), config={})
-
-        content = 'Use <think>example</think> in the document.'
-        result = requester._process_thinking_content(content, None, remove_think=True)
-        assert result == 'Use  in the document.'
-
-    def test_remove_initial_orphan_think_close(self):
-        """Test removing leading reasoning content when only </think> is visible"""
-        requester = litellmchat.LiteLLMRequester(ap=Mock(), config={})
-
-        content = 'hidden reasoning</think> Visible answer.'
-        result = requester._process_thinking_content(content, None, remove_think=True)
-        assert result == 'Visible answer.'
-
-    def test_remove_multiple_think_tags(self):
-        """Test removing multiple <think> blocks"""
-        requester = litellmchat.LiteLLMRequester(ap=Mock(), config={})
-
-        content = '<think>hidden</think> Keep <think>example</think>.'
-        result = requester._process_thinking_content(content, None, remove_think=True)
-        assert result == 'Keep .'
-
    def test_preserve_thinking_markers(self):
        """Test preserving thinking markers when remove_think=False"""
        requester = litellmchat.LiteLLMRequester(ap=Mock(), config={})
@@ -639,20 +491,6 @@ class TestProcessThinkingContent:
        assert 'CRETIRE_REASONING_BEGINk' in result
        assert 'The answer is 42.' in result

-    def test_preserve_reasoning_content_when_remove_think_false(self):
-        """Test showing separate reasoning_content when remove_think=False"""
-        requester = litellmchat.LiteLLMRequester(ap=Mock(), config={})
-
-        result = requester._process_thinking_content('The answer is 42.', 'Let me think...', remove_think=False)
-        assert result == '<think>\nLet me think...\n</think>\nThe answer is 42.'
-
-    def test_hide_reasoning_content_when_remove_think_true(self):
-        """Test hiding separate reasoning_content when remove_think=True"""
-        requester = litellmchat.LiteLLMRequester(ap=Mock(), config={})
-
-        result = requester._process_thinking_content('The answer is 42.', 'Let me think...', remove_think=True)
-        assert result == 'The answer is 42.'
-
    def test_empty_content(self):
        """Test empty content"""
        requester = litellmchat.LiteLLMRequester(ap=Mock(), config={})
@@ -1147,46 +985,6 @@ class TestInvokeRerank:
        assert results[0]['relevance_score'] == 1.0
        assert results[1]['relevance_score'] == 0.0

-    @pytest.mark.asyncio
-    @pytest.mark.parametrize(
-        ('model_extra_args', 'expected_url'),
-        [
-            ({'rerank_path': 'reranks'}, 'https://gateway.example.com/v1/reranks'),
-            ({'rerank_url': 'https://rerank.example.com/api/rerank'}, 'https://rerank.example.com/api/rerank'),
-        ],
-    )
-    async def test_invoke_rerank_openai_compatible_endpoint_override(self, model_extra_args, expected_url):
-        """Endpoint configuration controls routing and is not sent in the Cohere body."""
-        requester = litellmchat.LiteLLMRequester(
-            ap=Mock(),
-            config={
-                'base_url': 'https://gateway.example.com/v1/',
-                'custom_llm_provider': 'openai',
-            },
-        )
-        model = MockRuntimeRerankModel('Qwen3-Reranker-8B', 'test-api-key')
-        model.model_entity.extra_args = model_extra_args
-
-        mock_resp = Mock()
-        mock_resp.raise_for_status = Mock()
-        mock_resp.json = Mock(return_value={'results': [{'index': 0, 'relevance_score': 0.8}]})
-        mock_client = AsyncMock()
-        mock_client.post = AsyncMock(return_value=mock_resp)
-        mock_client.__aenter__ = AsyncMock(return_value=mock_client)
-        mock_client.__aexit__ = AsyncMock(return_value=False)
-
-        with patch('httpx.AsyncClient', return_value=mock_client):
-            await requester.invoke_rerank(model=model, query='query', documents=['document'])
-
-        assert mock_client.post.call_args.args[0] == expected_url
-        payload = mock_client.post.call_args.kwargs['json']
-        assert payload == {
-            'model': 'Qwen3-Reranker-8B',
-            'query': 'query',
-            'documents': ['document'],
-            'top_n': 1,
-        }
-

 class TestConvertMessages:
    """Test _convert_messages method"""
@@ -1,217 +0,0 @@
-from __future__ import annotations
-
-import json
-from types import SimpleNamespace
-from unittest.mock import AsyncMock, Mock
-
-import pytest
-
-import langbot_plugin.api.entities.builtin.pipeline.query as pipeline_query
-import langbot_plugin.api.entities.builtin.provider.message as provider_message
-import langbot_plugin.api.entities.builtin.provider.session as provider_session
-
-from langbot.pkg.provider.runners.localagent import LocalAgentRunner
-
-
-class RecordingProvider:
-    """Non-streaming provider that returns a tool-call on round 1 and plain text on round 2."""
-
-    def __init__(self):
-        self.requests: list[dict] = []
-
-    async def invoke_llm(self, query, model, messages, funcs, extra_args=None, remove_think=None):
-        self.requests.append(
-            {
-                'messages': list(messages),
-                'funcs': list(funcs),
-                'remove_think': remove_think,
-            }
-        )
-
-        if len(self.requests) == 1:
-            return provider_message.Message(
-                role='assistant',
-                content='Let me check that.',
-                tool_calls=[
-                    provider_message.ToolCall(
-                        id='call-1',
-                        type='function',
-                        function=provider_message.FunctionCall(
-                            name='exec',
-                            arguments=json.dumps({'command': "python -c 'print(42)'"}),
-                        ),
-                    )
-                ],
-            )
-
-        return provider_message.Message(
-            role='assistant',
-            content='The result is 42.',
-        )
-
-
-class RecordingStreamProvider:
-    """Streaming provider that returns a tool-call on round 1 and plain text on round 2."""
-
-    def __init__(self):
-        self.stream_requests: list[dict] = []
-
-    def invoke_llm_stream(self, query, model, messages, funcs, extra_args=None, remove_think=None):
-        self.stream_requests.append(
-            {
-                'messages': list(messages),
-                'funcs': list(funcs),
-                'remove_think': remove_think,
-            }
-        )
-
-        async def _stream():
-            if len(self.stream_requests) == 1:
-                yield provider_message.MessageChunk(
-                    role='assistant',
-                    content='Let me check that.',
-                    tool_calls=[
-                        provider_message.ToolCall(
-                            id='call-1',
-                            type='function',
-                            function=provider_message.FunctionCall(
-                                name='exec',
-                                arguments=json.dumps({'command': "python -c 'print(42)'"}),
-                            ),
-                        )
-                    ],
-                    is_final=True,
-                )
-                return
-
-            yield provider_message.MessageChunk(
-                role='assistant',
-                content='The result is 42.',
-                is_final=True,
-            )
-
-        return _stream()
-
-
-def make_query() -> pipeline_query.Query:
-    adapter = AsyncMock()
-    adapter.is_stream_output_supported = AsyncMock(return_value=False)
-
-    return pipeline_query.Query.model_construct(
-        query_id='no-dup-query',
-        launcher_type=provider_session.LauncherTypes.PERSON,
-        launcher_id=12345,
-        sender_id=12345,
-        message_chain=[],
-        message_event=None,
-        adapter=adapter,
-        pipeline_uuid='pipeline-uuid',
-        bot_uuid='bot-uuid',
-        pipeline_config={
-            'ai': {
-                'runner': {'runner': 'local-agent'},
-                'local-agent': {'model': {'primary': 'test-model-uuid', 'fallbacks': []}, 'prompt': 'test-prompt'},
-            },
-            'output': {'misc': {'remove-think': False}},
-        },
-        prompt=SimpleNamespace(messages=[]),
-        messages=[],
-        user_message=provider_message.Message(
-            role='user',
-            content='What is the answer?',
-        ),
-        use_funcs=[SimpleNamespace(name='exec')],
-        use_llm_model_uuid='test-model-uuid',
-        variables={},
-    )
-
-
-def _make_app(provider) -> SimpleNamespace:
-    model = SimpleNamespace(
-        provider=provider,
-        model_entity=SimpleNamespace(
-            uuid='test-model-uuid',
-            name='test-model',
-            abilities=['func_call'],
-            extra_args={},
-        ),
-    )
-    return SimpleNamespace(
-        logger=Mock(),
-        model_mgr=SimpleNamespace(get_model_by_uuid=AsyncMock(return_value=model)),
-        tool_mgr=SimpleNamespace(
-            execute_func_call=AsyncMock(
-                return_value={
-                    'session_id': 'no-dup-query',
-                    'backend': 'podman',
-                    'status': 'completed',
-                    'ok': True,
-                    'exit_code': 0,
-                    'stdout': '42',
-                    'stderr': '',
-                    'duration_ms': 10,
-                }
-            )
-        ),
-        rag_mgr=SimpleNamespace(),
-        box_service=SimpleNamespace(
-            get_system_guidance=Mock(return_value='sandbox guidance'),
-        ),
-        skill_mgr=SimpleNamespace(
-            get_skills_for_pipeline=AsyncMock(return_value=[]),
-            detect_skill_activation=AsyncMock(return_value=None),
-            build_activation_prompt=Mock(return_value=None),
-        ),
-    )
-
-
-@pytest.mark.asyncio
-async def test_localagent_non_streaming_no_duplicate():
-    """Non-streaming: round-2 content must not contain round-1 text."""
-    provider = RecordingProvider()
-    app = _make_app(provider)
-
-    runner = LocalAgentRunner(app, pipeline_config={})
-    query = make_query()
-
-    results = [message async for message in runner.run(query)]
-
-    # Expect: assistant (tool call) -> tool -> assistant (final answer)
-    assert [message.role for message in results] == ['assistant', 'tool', 'assistant']
-
-    final_message = results[-1]
-    assert final_message.content == 'The result is 42.'
-    assert 'Let me check that.' not in final_message.content
-
-
-@pytest.mark.asyncio
-async def test_localagent_streaming_no_duplicate():
-    """Streaming: round-2 content must not be re-seeded with round-1 text.
-
-    Regression test for the bug where _StreamAccumulator was initialized with
-    initial_content=first_content, causing every subsequent round to repeat
-    the entire opening line.
-    """
-    provider = RecordingStreamProvider()
-    app = _make_app(provider)
-
-    adapter = AsyncMock()
-    adapter.is_stream_output_supported = AsyncMock(return_value=True)
-
-    query = make_query()
-    query.adapter = adapter
-
-    runner = LocalAgentRunner(app, pipeline_config={})
-
-    results = [message async for message in runner.run(query)]
-
-    # All yielded messages should be MessageChunk in streaming mode
-    assert all(isinstance(message, provider_message.MessageChunk) for message in results)
-
-    # The last assistant chunk is the final answer for round 2
-    assistant_chunks = [m for m in results if m.role == 'assistant']
-    assert len(assistant_chunks) >= 2
-
-    final_chunk = assistant_chunks[-1]
-    assert final_chunk.content == 'The result is 42.'
-    assert 'Let me check that.' not in final_chunk.content
@@ -163,51 +163,6 @@ def test_stream_accumulator_merges_fragmented_tool_call_arguments():
    assert final_msg.tool_calls[0].function.arguments == '{"command":"pwd"}'


-def test_stream_accumulator_strips_leading_think_from_tool_round_content():
-    accumulator = _StreamAccumulator(
-        msg_sequence=3,
-        initial_content='I will search for LangBot.',
-        remove_think=True,
-    )
-
-    assert accumulator.add(provider_message.MessageChunk(role='assistant', content='<thi')) is None
-    assert accumulator.add(provider_message.MessageChunk(role='assistant', content='nk>hidden')) is None
-    emitted = accumulator.add(
-        provider_message.MessageChunk(
-            role='assistant',
-            content=' reasoning</think>Here is the answer.',
-            is_final=True,
-        )
-    )
-
-    assert emitted is not None
-    assert emitted.content == 'I will search for LangBot.Here is the answer.'
-    assert '<think>' not in emitted.content
-    assert 'hidden reasoning' not in emitted.content
-
-
-def test_stream_accumulator_strips_initial_orphan_think_close_from_tool_round_content():
-    accumulator = _StreamAccumulator(
-        msg_sequence=3,
-        initial_content='I will search for LangBot.',
-        remove_think=True,
-    )
-
-    assert accumulator.add(provider_message.MessageChunk(role='assistant', content='hidden reasoning')) is None
-    emitted = accumulator.add(
-        provider_message.MessageChunk(
-            role='assistant',
-            content=' still hidden</think>Here is the answer.',
-            is_final=True,
-        )
-    )
-
-    assert emitted is not None
-    assert emitted.content == 'I will search for LangBot.Here is the answer.'
-    assert '</think>' not in emitted.content
-    assert 'hidden reasoning' not in emitted.content
-
-
@pytest.mark.asyncio
 async def test_localagent_uses_exec_for_exact_calculation():
    provider = RecordingProvider()
@@ -417,7 +417,7 @@ class TestBuildBoxSessionPayload:
        payload = s._build_box_session_payload('session-123')
        assert payload['image'] == 'node:20'
        assert payload['cpus'] == 2.0
-        assert payload['memory_mb'] == 1024
+        assert payload["memory_mb"] == 1024
        assert payload['pids_limit'] == 256

    def test_none_fields_excluded(self, mcp_module):
@@ -680,55 +680,14 @@ class TestGetRuntimeInfoDict:
        # ... but are isolated by distinct process_ids within that session.
        assert transient._box_stdio_runtime.process_id != live._box_stdio_runtime.process_id

-    def test_different_resource_profiles_use_different_box_sessions(self, mcp_module):
-        ap = _make_ap()
-        ap.box_service.available = True
-        default = _make_session(
-            mcp_module,
-            {
-                'name': 'default',
-                'uuid': 'default-uuid',
-                'mode': 'stdio',
-                'command': 'uvx',
-                'args': ['mcp-server-time'],
-            },
-            ap=ap,
-        )
-        constrained = _make_session(
-            mcp_module,
-            {
-                'name': 'constrained',
-                'uuid': 'constrained-uuid',
-                'mode': 'stdio',
-                'command': 'uvx',
-                'args': ['mcp-server-time'],
-                'box': {'memory_mb': 2048},
-            },
-            ap=ap,
-        )
-
-        assert default._build_box_session_id() == 'mcp-shared'
-        assert constrained._build_box_session_id().startswith('mcp-shared-')
-        assert constrained._build_box_session_id() != default._build_box_session_id()
-
-        writable = _make_session(
-            mcp_module,
-            {
-                'name': 'writable',
-                'uuid': 'writable-uuid',
-                'mode': 'stdio',
-                'command': 'uvx',
-                'args': ['mcp-server-time'],
-                'box': {'host_path_mode': 'rw'},
-            },
-            ap=ap,
-        )
-        assert writable._build_box_session_id() != default._build_box_session_id()
-
-    def test_stdio_session_waits_for_enabled_box_when_temporarily_unavailable(self, mcp_module):
+    def test_stdio_session_refuses_when_box_unavailable(self, mcp_module):
+        """Policy: when Box is configured but unavailable (disabled in config
+        OR connection failed), stdio MCP servers are NOT treated as box-stdio.
+        ``_init_stdio_python_server`` will raise a clear refusal at start
+        time; until then, the runtime info simply omits box_session_id so the
+        UI can render the disabled state cleanly."""
        ap = _make_ap()
        ap.box_service.available = False
-        ap.box_service.enabled = True
        s = _make_session(
            mcp_module,
            {
@@ -741,125 +700,9 @@ class TestGetRuntimeInfoDict:
            ap=ap,
        )
        info = s.get_runtime_info_dict()
-        assert info['box_session_id'] == 'mcp-shared'
-        assert info['box_enabled'] is True
-
-    def test_stdio_session_refuses_when_box_is_disabled(self, mcp_module):
-        ap = _make_ap()
-        ap.box_service.available = False
-        ap.box_service.enabled = False
-        s = _make_session(
-            mcp_module,
-            {
-                'name': 'test',
-                'uuid': 'test-uuid',
-                'mode': 'stdio',
-                'command': 'python',
-                'args': [],
-            },
-            ap=ap,
-        )
-
-        info = s.get_runtime_info_dict()
-
        assert 'box_session_id' not in info
        assert 'box_enabled' not in info

-    @pytest.mark.asyncio
-    async def test_stdio_session_waits_until_box_reconnects(self, mcp_module, monkeypatch):
-        mcp_stdio_module = sys.modules['langbot.pkg.provider.tools.loaders.mcp_stdio']
-        ap = _make_ap()
-        ap.box_service.available = False
-        ap.box_service.enabled = True
-        session = _make_session(
-            mcp_module,
-            {
-                'name': 'test',
-                'uuid': 'test-uuid',
-                'mode': 'stdio',
-                'command': 'python',
-                'args': [],
-                'box': {'startup_timeout_sec': 1},
-            },
-            ap=ap,
-        )
-
-        async def reconnect_box(_delay):
-            ap.box_service.available = True
-
-        monkeypatch.setattr(mcp_stdio_module.asyncio, 'sleep', reconnect_box)
-
-        await session._box_stdio_runtime._wait_for_box_runtime()
-
-        assert ap.box_service.available is True
-
-    @pytest.mark.asyncio
-    async def test_enabled_box_timeout_does_not_exhaust_mcp_retry_budget(self, mcp_module, monkeypatch):
-        ap = _make_ap()
-        ap.box_service.available = False
-        ap.box_service.enabled = True
-        session = _make_session(
-            mcp_module,
-            {
-                'name': 'test',
-                'uuid': 'test-uuid',
-                'mode': 'stdio',
-                'command': 'python',
-                'args': [],
-            },
-            ap=ap,
-        )
-        attempts = 0
-
-        async def lifecycle():
-            nonlocal attempts
-            attempts += 1
-            if attempts == 1:
-                session.error_phase = mcp_module.MCPSessionErrorPhase.BOX_UNAVAILABLE
-                raise RuntimeError('Box runtime is not available after 1 seconds')
-
-        session._lifecycle_loop = lifecycle
-        sleep = AsyncMock()
-        monkeypatch.setattr(mcp_module.asyncio, 'sleep', sleep)
-
-        await session._lifecycle_loop_with_retry()
-
-        assert attempts == 2
-        assert session.retry_count == 0
-        sleep.assert_awaited_once_with(1)
-
-    @pytest.mark.asyncio
-    async def test_disabled_box_still_stops_mcp_retry_loop(self, mcp_module):
-        ap = _make_ap()
-        ap.box_service.available = False
-        ap.box_service.enabled = False
-        session = _make_session(
-            mcp_module,
-            {
-                'name': 'test',
-                'uuid': 'test-uuid',
-                'mode': 'stdio',
-                'command': 'python',
-                'args': [],
-            },
-            ap=ap,
-        )
-        attempts = 0
-
-        async def lifecycle():
-            nonlocal attempts
-            attempts += 1
-            session.error_phase = mcp_module.MCPSessionErrorPhase.BOX_UNAVAILABLE
-            raise RuntimeError('box_disabled_in_config')
-
-        session._lifecycle_loop = lifecycle
-
-        await session._lifecycle_loop_with_retry()
-
-        assert attempts == 1
-        assert session.status == mcp_module.MCPSessionStatus.ERROR
-        assert session.retry_count == 1
-
    def test_stdio_session_without_box_service_uses_local_stdio(self, mcp_module):
        ap = _make_ap()
        del ap.box_service
@@ -12,7 +12,7 @@ import pytest
 from aiohttp import web
 from mcp import types as mcp_types

-from langbot.pkg.provider.tools.loaders.mcp import MCPToolCallTimeoutError, RuntimeMCPSession
+from langbot.pkg.provider.tools.loaders.mcp import RuntimeMCPSession


 class _TransportProbe:
@@ -65,25 +65,6 @@ class _TransportProbe:
                            },
                        }
                    )
-                if method == 'tools/call':
-                    tool_name = message.get('params', {}).get('name')
-                    if tool_name == 'hang':
-                        return web.Response(status=202)
-                    return web.json_response(
-                        {
-                            'jsonrpc': '2.0',
-                            'id': message['id'],
-                            'result': {
-                                'content': [
-                                    {
-                                        'type': 'text',
-                                        'text': 'healthy',
-                                    }
-                                ],
-                                'isError': False,
-                            },
-                        }
-                    )
                return web.Response(status=202)
            return web.Response(status=self.streamable_status)

@@ -145,22 +126,11 @@ async def _transport_server(streamable_status: int | None):
        await runner.cleanup()


-def _session(
-    url: str,
-    *,
-    timeout: float = 2,
-    tool_call_timeout_sec: float = 300,
-) -> RuntimeMCPSession:
+def _session(url: str, *, timeout: float = 2) -> RuntimeMCPSession:
    app = cast(Any, SimpleNamespace(logger=Mock()))
    return RuntimeMCPSession(
        'remote-transport-test',
-        {
-            'uuid': 'srv-1',
-            'mode': 'remote',
-            'url': url,
-            'timeout': timeout,
-            'tool_call_timeout_sec': tool_call_timeout_sec,
-        },
+        {'uuid': 'srv-1', 'mode': 'remote', 'url': url, 'timeout': timeout},
        True,
        app,
    )
@@ -194,24 +164,6 @@ async def test_remote_transport_real_streamable_http_success_keeps_session_usabl
            await _close_session(session)


-@pytest.mark.asyncio
-async def test_remote_transport_tool_timeout_does_not_poison_session():
-    async with _transport_server(200) as (probe, url):
-        session = _session(url, tool_call_timeout_sec=0.05)
-        try:
-            await session._init_remote_server()
-
-            with pytest.raises(MCPToolCallTimeoutError, match='timed out after 0.05 seconds'):
-                await session.invoke_mcp_tool('hang', {})
-
-            result = await session.invoke_mcp_tool('health_check', {})
-
-            assert result[0].text == 'healthy'
-            assert probe.streamable_messages.count('tools/call') == 2
-        finally:
-            await _close_session(session)
-
-
@pytest.mark.asyncio
@pytest.mark.parametrize('status_code', [400, 404, 405])
 async def test_remote_transport_real_streamable_http_error_falls_back_to_legacy_sse(status_code: int):
@@ -1,25 +1,20 @@
 from __future__ import annotations

-import asyncio
 import base64
-from datetime import timedelta
 from types import SimpleNamespace
 from unittest.mock import AsyncMock, Mock

 import httpx
 import pytest
 from mcp import types as mcp_types
-from mcp.shared.exceptions import McpError

 from langbot.pkg.provider.tools.loaders.mcp import (
    MCP_RESOURCE_CONTEXT_QUERY_KEY,
    MCP_RESOURCE_TRACE_QUERY_KEY,
-    MCP_TOOL_CALL_TIMEOUT_DEFAULT_SECONDS,
    MCP_TOOL_LIST_RESOURCES,
    MCP_TOOL_READ_RESOURCE,
    MCPLoader,
    MCPSessionStatus,
-    MCPToolCallTimeoutError,
    RuntimeMCPSession,
 )
 from langbot.pkg.telemetry import features as telemetry_features
@@ -66,111 +61,6 @@ def _http_status_error(status_code: int) -> httpx.HTTPStatusError:
    return httpx.HTTPStatusError(f'HTTP {status_code}', request=request, response=response)


-def _tool_result(text: str = 'ok') -> mcp_types.CallToolResult:
-    return mcp_types.CallToolResult(
-        content=[mcp_types.TextContent(type='text', text=text)],
-        isError=False,
-    )
-
-
-@pytest.mark.asyncio
-async def test_invoke_mcp_tool_uses_configurable_request_timeout():
-    session = RuntimeMCPSession(
-        'slow-tools',
-        {
-            'uuid': 'srv-1',
-            'mode': 'remote',
-            'tool_call_timeout_sec': 900,
-        },
-        True,
-        _app(),
-    )
-    session.session = SimpleNamespace(call_tool=AsyncMock(return_value=_tool_result()))
-
-    result = await session.invoke_mcp_tool('render_video', {'quality': 'high'})
-
-    assert result[0].text == 'ok'
-    session.session.call_tool.assert_awaited_once_with(
-        'render_video',
-        {'quality': 'high'},
-        read_timeout_seconds=timedelta(seconds=900),
-    )
-
-
-@pytest.mark.asyncio
-async def test_invoke_mcp_tool_zero_timeout_disables_request_deadline():
-    session = RuntimeMCPSession(
-        'unbounded-tools',
-        {
-            'uuid': 'srv-1',
-            'mode': 'remote',
-            'tool_call_timeout_sec': 0,
-        },
-        True,
-        _app(),
-    )
-    session.session = SimpleNamespace(call_tool=AsyncMock(return_value=_tool_result()))
-
-    await session.invoke_mcp_tool('long_job', {})
-
-    session.session.call_tool.assert_awaited_once_with(
-        'long_job',
-        {},
-        read_timeout_seconds=None,
-    )
-
-
-@pytest.mark.asyncio
-async def test_invoke_mcp_tool_timeout_is_not_retried_and_session_remains_usable():
-    session = RuntimeMCPSession(
-        'recoverable-tools',
-        {
-            'uuid': 'srv-1',
-            'mode': 'remote',
-            'tool_call_timeout_sec': 5,
-        },
-        True,
-        _app(),
-    )
-    timeout = McpError(
-        mcp_types.ErrorData(
-            code=httpx.codes.REQUEST_TIMEOUT,
-            message='Timed out while waiting for response to ClientRequest. Waited 5 seconds.',
-        )
-    )
-    call_tool = AsyncMock(side_effect=[timeout, _tool_result('recovered')])
-    session.session = SimpleNamespace(call_tool=call_tool)
-
-    with pytest.raises(
-        MCPToolCallTimeoutError,
-        match="MCP tool 'long_job' on server 'recoverable-tools' timed out after 5 seconds",
-    ):
-        await session.invoke_mcp_tool('long_job', {})
-
-    assert call_tool.await_count == 1
-    second_result = await session.invoke_mcp_tool('health_check', {})
-    assert second_result[0].text == 'recovered'
-    assert call_tool.await_count == 2
-
-
-@pytest.mark.parametrize('invalid_timeout', [-1, float('inf'), 1e300, True, 'not-a-number'])
-def test_invalid_tool_call_timeout_falls_back_to_default(invalid_timeout):
-    ap = _app()
-    session = RuntimeMCPSession(
-        'invalid-timeout',
-        {
-            'uuid': 'srv-1',
-            'mode': 'remote',
-            'tool_call_timeout_sec': invalid_timeout,
-        },
-        True,
-        ap,
-    )
-
-    assert session.tool_call_timeout_sec == MCP_TOOL_CALL_TIMEOUT_DEFAULT_SECONDS
-    ap.logger.warning.assert_called_once()
-
-
@pytest.mark.asyncio
 async def test_remote_transport_falls_back_to_sse_for_compatible_http_status_in_exception_group():
    session = RuntimeMCPSession(
@@ -431,42 +321,3 @@ async def test_build_resource_context_for_query_uses_only_bound_attached_text_re
    assert query.variables[MCP_RESOURCE_CONTEXT_QUERY_KEY]['resource_count'] == 1
    docs.session.read_resource.assert_awaited_once()
    other.session.read_resource.assert_not_called()
-
-
-@pytest.mark.asyncio
-async def test_mcp_loader_shutdown_cancels_startup_tasks_and_closes_sessions_concurrently():
-    loader = MCPLoader(_app())
-    hosted_cancelled = asyncio.Event()
-
-    async def pending_host():
-        try:
-            await asyncio.Event().wait()
-        finally:
-            hosted_cancelled.set()
-
-    hosted_task = asyncio.create_task(pending_host())
-    await asyncio.sleep(0)
-    loader._hosted_mcp_tasks = [hosted_task]
-
-    started: set[str] = set()
-    all_started = asyncio.Event()
-
-    class Session:
-        def __init__(self, name: str):
-            self.name = name
-
-        async def shutdown(self):
-            started.add(self.name)
-            if len(started) == 2:
-                all_started.set()
-            await all_started.wait()
-
-    loader.sessions = {'one': Session('one'), 'two': Session('two')}
-
-    await asyncio.wait_for(loader.shutdown(), timeout=1)
-
-    assert hosted_cancelled.is_set()
-    assert hosted_task.cancelled()
-    assert started == {'one', 'two'}
-    assert loader._hosted_mcp_tasks == []
-    assert loader.sessions == {}
@@ -8,8 +8,7 @@ and error handling without calling real LLM APIs.
 from __future__ import annotations

 import pytest
-from types import SimpleNamespace
-from unittest.mock import AsyncMock, Mock
+from unittest.mock import Mock

 from langbot.pkg.provider.modelmgr.modelmgr import ModelManager
 from langbot.pkg.provider.modelmgr import requester
@@ -63,48 +62,6 @@ async def test_model_manager_skips_space_sync_when_disabled(mock_app_for_modelmg
    app.space_service.get_models.assert_not_called()


-@pytest.mark.asyncio
-async def test_sync_new_models_from_space_creates_rerank_models(mock_app_for_modelmgr):
-    """Space rerank entries are discovered and persisted under the shared provider."""
-    app = mock_app_for_modelmgr
-    provider = persistence_model.ModelProvider(
-        uuid='space-provider',
-        name='LangBot Space',
-        requester='space-chat-completions',
-        base_url='https://api.langbot.cloud/v1',
-        api_keys=['space-key'],
-    )
-    app.persistence_mgr.execute_async = AsyncMock(return_value=_make_mock_result([provider], first_item=provider))
-    app.space_service.get_models = AsyncMock(
-        return_value=[
-            SimpleNamespace(
-                uuid='rerank-model-uuid',
-                model_id='Qwen3-Reranker-8B',
-                category='rerank',
-                featured_order=10,
-            )
-        ]
-    )
-    app.llm_model_service.get_llm_models = AsyncMock(return_value=[])
-    app.embedding_models_service.get_embedding_models = AsyncMock(return_value=[])
-    app.rerank_models_service = AsyncMock()
-    app.rerank_models_service.get_rerank_models = AsyncMock(return_value=[])
-
-    model_mgr = ModelManager(app)
-    await model_mgr.sync_new_models_from_space()
-
-    app.rerank_models_service.create_rerank_model.assert_awaited_once_with(
-        {
-            'uuid': 'rerank-model-uuid',
-            'name': 'Qwen3-Reranker-8B',
-            'provider_uuid': 'space-provider',
-            'extra_args': {},
-            'prefered_ranking': 10,
-        },
-        preserve_uuid=True,
-    )
-
-
 # ============================================================================
 # Model Loading Tests
 # ============================================================================
@@ -458,25 +458,6 @@ class TestSkillToolLoader:
        assert await loader.has_tool('activate') is True
        assert await loader.has_tool('register_skill') is True

-    @pytest.mark.asyncio
-    async def test_tools_reappear_after_box_backend_recovers(self):
-        from langbot.pkg.provider.tools.loaders.skill_authoring import SkillToolLoader
-
-        ap = _make_ap()
-        ap.skill_mgr = SimpleNamespace(skills={'demo': _make_skill_data(name='demo')})
-        ap.box_service = SimpleNamespace(
-            available=False,
-            get_status=AsyncMock(return_value={'backend': {'available': True}}),
-        )
-
-        loader = SkillToolLoader(ap)
-        await loader.initialize()
-        assert await loader.get_tools() == []
-
-        ap.box_service.available = True
-
-        assert sorted(tool.name for tool in await loader.get_tools()) == ['activate', 'register_skill']
-

 class TestNativeToolLoaderSkillPaths:
    @pytest.mark.asyncio
@@ -151,21 +151,6 @@ async def test_native_tool_loader_exposes_all_tools_when_box_available():
        assert await loader.has_tool(tool_name) is True


-@pytest.mark.asyncio
-async def test_native_tool_loader_refreshes_after_box_recovers():
-    box_service = SimpleNamespace(
-        available=False,
-        get_status=AsyncMock(return_value={'backend': {'available': True}}),
-    )
-    loader = NativeToolLoader(SimpleNamespace(box_service=box_service, logger=Mock()))
-    await loader.initialize()
-    assert await loader.get_tools() == []
-
-    box_service.available = True
-
-    assert [tool.name for tool in await loader.get_tools()] == ['exec', 'read', 'write', 'edit', 'glob', 'grep']
-
-
 # ── read/write/edit file tool tests ─────────────────────────────


@@ -357,13 +357,10 @@ class TestCredentialsBuild:
            cred_calls.append(kwargs)
            return ('CRED', kwargs)

-        # These names are absent when the optional valkey-glide dependency is
-        # unavailable (for example, on Windows), so allow the test doubles to
-        # create them on the module.
-        monkeypatch.setattr(mod, 'GlideClient', _FakeClient, raising=False)
-        monkeypatch.setattr(mod, 'ServerCredentials', _fake_credentials, raising=False)
-        monkeypatch.setattr(mod, 'GlideClientConfiguration', lambda **kw: kw, raising=False)
-        monkeypatch.setattr(mod, 'NodeAddress', lambda *a, **k: ('node', a, k), raising=False)
+        monkeypatch.setattr(mod, 'GlideClient', _FakeClient)
+        monkeypatch.setattr(mod, 'ServerCredentials', _fake_credentials)
+        monkeypatch.setattr(mod, 'GlideClientConfiguration', lambda **kw: kw)
+        monkeypatch.setattr(mod, 'NodeAddress', lambda *a, **k: ('node', a, k))
        return backend, created, cred_calls, warnings

    async def test_username_without_password_fails_closed(self, monkeypatch):
@@ -2008,7 +2008,7 @@ wheels = [

 [[package]]
 name = "langbot"
-version = "4.10.6"
+version = "4.10.5"
 source = { editable = "." }
 dependencies = [
    { name = "aiocqhttp" },
@@ -2084,7 +2084,7 @@ dependencies = [
    { name = "tiktoken" },
    { name = "urllib3" },
    { name = "uv" },
-    { name = "valkey-glide", marker = "sys_platform != 'win32'" },
+    { name = "valkey-glide" },
    { name = "websockets" },
 ]

@@ -2124,7 +2124,7 @@ requires-dist = [
    { name = "ebooklib", specifier = ">=0.18" },
    { name = "gewechat-client", specifier = ">=0.1.5" },
    { name = "html2text", specifier = ">=2024.2.26" },
-    { name = "langbot-plugin", specifier = "==0.4.17" },
+    { name = "langbot-plugin", specifier = "==0.4.13" },
    { name = "langchain", specifier = ">=1.3.9" },
    { name = "langchain-core", specifier = ">=1.3.3" },
    { name = "langchain-text-splitters", specifier = ">=1.1.2" },
@@ -2173,7 +2173,7 @@ requires-dist = [
    { name = "tiktoken", specifier = ">=0.9.0" },
    { name = "urllib3", specifier = ">=2.7.0" },
    { name = "uv", specifier = ">=0.11.15" },
-    { name = "valkey-glide", marker = "sys_platform != 'win32'", specifier = ">=2.4.1,<3.0.0" },
+    { name = "valkey-glide", specifier = ">=2.4.1,<3.0.0" },
    { name = "websockets", specifier = ">=15.0.1" },
 ]

@@ -2189,7 +2189,7 @@ dev = [

 [[package]]
 name = "langbot-plugin"
-version = "0.4.17"
+version = "0.4.13"
 source = { registry = "https://pypi.org/simple" }
 dependencies = [
    { name = "aiofiles" },
@@ -2210,9 +2210,9 @@ dependencies = [
    { name = "watchdog" },
    { name = "websockets" },
 ]
-sdist = { url = "https://files.pythonhosted.org/packages/3a/96/336d7ac97ff5a9c413e7aaf0c329a7047ec91dec5fa4f9e0b1d0abb57d0e/langbot_plugin-0.4.17.tar.gz", hash = "sha256:b1539444f16568c0b3244f68c498ac9498c56e2c20f12e220e6ee735b8d6b5c2", size = 347080, upload-time = "2026-07-23T09:23:55.941Z" }
+sdist = { url = "https://files.pythonhosted.org/packages/40/a6/1eaf77c3b81e9de3390c504c5f627dc41f43bff6df9aff0e1e31d796b6f0/langbot_plugin-0.4.13.tar.gz", hash = "sha256:f936340e67679c21f1e7e7f1447339f31a0a2c965db060ecfbd9d0c51bb0d6fe", size = 334887, upload-time = "2026-07-04T05:38:59.942Z" }
 wheels = [
-    { url = "https://files.pythonhosted.org/packages/e1/98/ac372a238e59894ac54189cb220a3fb75da48f319881eeb49d583712549a/langbot_plugin-0.4.17-py3-none-any.whl", hash = "sha256:6fd6c9d7e0583f04659f023a233c4201cb4c450d1ef696e79fa1e338933ee156", size = 229881, upload-time = "2026-07-23T09:23:54.582Z" },
+    { url = "https://files.pythonhosted.org/packages/e9/bf/fc9671a7afbd933440c38403c84d918c1022fdeed16e22a6ab3b2aec83ff/langbot_plugin-0.4.13-py3-none-any.whl", hash = "sha256:9d45ebc7a7ee0413d6db9baa009fcbf0ad07e2e1753a6f0a27f37b8b665cd1ee", size = 221884, upload-time = "2026-07-04T05:38:58.525Z" },
 ]

 [[package]]
@@ -5991,9 +5991,9 @@ name = "valkey-glide"
 version = "2.4.1"
 source = { registry = "https://pypi.org/simple" }
 dependencies = [
-    { name = "anyio", marker = "sys_platform != 'win32'" },
-    { name = "protobuf", marker = "sys_platform != 'win32'" },
-    { name = "sniffio", marker = "sys_platform != 'win32'" },
+    { name = "anyio" },
+    { name = "protobuf" },
+    { name = "sniffio" },
 ]
 sdist = { url = "https://files.pythonhosted.org/packages/72/a2/582b34c6acc8dc857c537f6007459cba48dfa0dc404789a657e5c1a998c0/valkey_glide-2.4.1.tar.gz", hash = "sha256:f1155d84156d11b90488aa67e90102f0bf98a45314f5b99308ac9074c05f7241", size = 898030, upload-time = "2026-05-28T21:41:55.881Z" }
 wheels = [
@@ -6,7 +6,6 @@
    "dev": "vite",
    "build": "tsc && vite build",
    "preview": "vite preview",
-    "test:unit": "node --test tests/unit/*.test.mjs",
    "test:e2e": "playwright test",
    "lint": "eslint .",
    "format": "prettier --write ."
@@ -32,7 +31,6 @@
    "@dnd-kit/core": "^6.3.1",
    "@dnd-kit/sortable": "^10.0.0",
    "@dnd-kit/utilities": "^3.2.2",
-    "@fingerprintjs/fingerprintjs": "^4.6.2",
    "@hookform/resolvers": "^5.0.1",
    "@radix-ui/react-alert-dialog": "^1.1.15",
    "@radix-ui/react-avatar": "^1.1.11",
@@ -24,9 +24,6 @@ dependencies:
  '@dnd-kit/utilities':
    specifier: ^3.2.2
    version: 3.2.2(react@19.2.1)
-  '@fingerprintjs/fingerprintjs':
-    specifier: ^4.6.2
-    version: 4.6.2
  '@hookform/resolvers':
    specifier: ^5.0.1
    version: 5.2.2(react-hook-form@7.71.1)
@@ -425,12 +422,6 @@ packages:
      levn: 0.4.1
    dev: true

-  /@fingerprintjs/fingerprintjs@4.6.2:
-    resolution: {integrity: sha512-g8mXuqcFKbgH2CZKwPfVtsUJDHyvcgIABQI7Y0tzWEFXpGxJaXuAuzlifT2oTakjDBLTK4Gaa9/5PERDhqUjtw==}
-    dependencies:
-      tslib: 2.8.1
-    dev: false
-
  /@floating-ui/core@1.7.4:
    resolution: {integrity: sha512-C3HlIdsBxszvm5McXlB8PeOEWfBhcGBTZGkGlWc2U0KFY5IwG5OQEuQ8rq52DZmcHDlPLd+YFBK+cZcytwIFWg==}
    dependencies:
@@ -15,7 +15,6 @@ import {
  FormMessage,
 } from '@/components/ui/form';
 import DynamicFormItemComponent from '@/app/home/components/dynamic-form/DynamicFormItemComponent';
-import { normalizeDynamicFormValuesForSave } from '@/app/home/components/dynamic-form/DynamicFormSaveValues';
 import QrCodeLoginDialog, {
  QrLoginPlatform,
 } from '@/app/home/components/qrcode-login/QrCodeLoginDialog';
@@ -632,9 +631,12 @@ export default function DynamicFormComponent({
    // even if the user saves without modifying any field.
    // form.watch(callback) only fires on subsequent changes, not on mount.
    const formValues = form.getValues();
-    const initialFinalValues = normalizeDynamicFormValuesForSave(
-      editableValueSpecs,
-      formValues as Record<string, unknown>,
+    const initialFinalValues = editableValueSpecs.reduce(
+      (acc, item) => {
+        acc[item.name] = formValues[item.name] ?? item.default;
+        return acc;
+      },
+      {} as Record<string, object>,
    );
    onSubmitRef.current?.(initialFinalValues);

@@ -649,9 +651,12 @@ export default function DynamicFormComponent({

    const subscription = form.watch(() => {
      const formValues = form.getValues();
-      const finalValues = normalizeDynamicFormValuesForSave(
-        editableValueSpecs,
-        formValues as Record<string, unknown>,
+      const finalValues = editableValueSpecs.reduce(
+        (acc, item) => {
+          acc[item.name] = formValues[item.name] ?? item.default;
+          return acc;
+        },
+        {} as Record<string, object>,
      );
      onSubmitRef.current?.(finalValues);
      previousInitialValues.current = finalValues as Record<string, object>;
@@ -1,25 +0,0 @@
-import type { IDynamicFormItemSchema } from '@/app/infra/entities/form/dynamic';
-
-export type DynamicFormSaveValueSpec = Pick<
-  IDynamicFormItemSchema,
-  'default' | 'name' | 'type'
->;
-
-/**
- * Build the value snapshot emitted to parent forms for persistence.
- * Only single-line string fields trim surrounding whitespace; multiline text
- * and every other dynamic form field type preserve their original values.
- */
-export function normalizeDynamicFormValuesForSave(
-  specs: readonly DynamicFormSaveValueSpec[],
-  formValues: Record<string, unknown>,
-): Record<string, unknown> {
-  return specs.reduce<Record<string, unknown>>((values, spec) => {
-    const value = formValues[spec.name] ?? spec.default;
-    values[spec.name] =
-      spec.type === 'string' && typeof value === 'string'
-        ? value.trim()
-        : value;
-    return values;
-  }, {});
-}
@@ -12,7 +12,6 @@ import {
 } from '@/components/ui/form';
 import { IDynamicFormItemSchema } from '@/app/infra/entities/form/dynamic';
 import DynamicFormItemComponent from '@/app/home/components/dynamic-form/DynamicFormItemComponent';
-import { normalizeDynamicFormValuesForSave } from '@/app/home/components/dynamic-form/DynamicFormSaveValues';
 import { extractI18nObject } from '@/i18n/I18nProvider';

 /**
@@ -151,9 +150,12 @@ export default function N8nAuthFormComponent({
    // Emit initial form values on mount so the parent form's
    // initializedStagesRef registers this stage (matches DynamicFormComponent).
    const formValues = form.getValues();
-    const initialFinalValues = normalizeDynamicFormValuesForSave(
-      itemConfigList,
-      formValues as Record<string, unknown>,
+    const initialFinalValues = itemConfigList.reduce(
+      (acc, item) => {
+        acc[item.name] = formValues[item.name] ?? item.default;
+        return acc;
+      },
+      {} as Record<string, string>,
    );
    onSubmitRef.current?.(initialFinalValues);
    previousInitialValues.current = initialFinalValues as Record<
@@ -169,9 +171,12 @@ export default function N8nAuthFormComponent({

      // 获取完整的表单值，确保包含所有默认值
      const formValues = form.getValues();
-      const finalValues = normalizeDynamicFormValuesForSave(
-        itemConfigList,
-        formValues as Record<string, unknown>,
+      const finalValues = itemConfigList.reduce(
+        (acc, item) => {
+          acc[item.name] = formValues[item.name] ?? item.default;
+          return acc;
+        },
+        {} as Record<string, string>,
      );

      onSubmitRef.current?.(finalValues);
@@ -435,10 +435,6 @@ const getFormSchema = (t: TFunction) =>
        .number({ invalid_type_error: t('mcp.timeoutMustBeNumber') })
        .positive({ message: t('mcp.timeoutMustBePositive') })
        .default(30),
-      tool_call_timeout_sec: z
-        .number({ invalid_type_error: t('mcp.timeoutMustBeNumber') })
-        .nonnegative({ message: t('mcp.timeoutNonNegative') })
-        .default(300),
      ssereadtimeout: z
        .number({ invalid_type_error: t('mcp.sseTimeoutMustBeNumber') })
        .positive({ message: t('mcp.timeoutMustBePositive') })
@@ -478,7 +474,6 @@ const getFormSchema = (t: TFunction) =>

 type FormValues = z.infer<ReturnType<typeof getFormSchema>> & {
  timeout: number;
-  tool_call_timeout_sec: number;
  ssereadtimeout: number;
 };

@@ -540,7 +535,6 @@ const MCPForm = forwardRef<MCPFormHandle, MCPFormProps>(function MCPForm(
      command: '',
      args: [],
      timeout: 30,
-      tool_call_timeout_sec: 300,
      ssereadtimeout: 300,
      extra_args: [],
      ...initialDraftRef.current,
@@ -615,7 +609,6 @@ const MCPForm = forwardRef<MCPFormHandle, MCPFormProps>(function MCPForm(
        command: '',
        args: [],
        timeout: 30,
-        tool_call_timeout_sec: 300,
        ssereadtimeout: 300,
        extra_args: [],
        ...initialDraftRef.current,
@@ -694,16 +687,10 @@ const MCPForm = forwardRef<MCPFormHandle, MCPFormProps>(function MCPForm(
        command: '',
        args: [],
        timeout: 30,
-        tool_call_timeout_sec: 300,
        ssereadtimeout: 300,
        extra_args: [],
      };

-      if (typeof server.extra_args.tool_call_timeout_sec === 'number') {
-        formValues.tool_call_timeout_sec =
-          server.extra_args.tool_call_timeout_sec;
-      }
-
      let newExtraArgs: {
        key: string;
        type: 'string' | 'number' | 'boolean';
@@ -783,7 +770,6 @@ const MCPForm = forwardRef<MCPFormHandle, MCPFormProps>(function MCPForm(
            url: value.url!,
            headers,
            timeout: value.timeout,
-            tool_call_timeout_sec: value.tool_call_timeout_sec,
          },
        };
      } else {
@@ -800,7 +786,6 @@ const MCPForm = forwardRef<MCPFormHandle, MCPFormProps>(function MCPForm(
            command: value.command!,
            args: value.args?.map((arg) => arg.value) || [],
            env,
-            tool_call_timeout_sec: value.tool_call_timeout_sec,
          },
        };
      }
@@ -850,7 +835,6 @@ const MCPForm = forwardRef<MCPFormHandle, MCPFormProps>(function MCPForm(
        extraArgsData = {
          url: form.getValues('url')!,
          timeout: form.getValues('timeout'),
-          tool_call_timeout_sec: form.getValues('tool_call_timeout_sec'),
          headers: Object.fromEntries(
            formExtraArgs.map((arg) => [arg.key, arg.value]),
          ),
@@ -862,7 +846,6 @@ const MCPForm = forwardRef<MCPFormHandle, MCPFormProps>(function MCPForm(
          env: Object.fromEntries(
            formExtraArgs.map((arg) => [arg.key, arg.value]),
          ),
-          tool_call_timeout_sec: form.getValues('tool_call_timeout_sec'),
        };
      }

@@ -1064,30 +1047,6 @@ const MCPForm = forwardRef<MCPFormHandle, MCPFormProps>(function MCPForm(
          )}
        />

-        <FormField
-          control={form.control}
-          name="tool_call_timeout_sec"
-          render={({ field }) => (
-            <FormItem>
-              <FormLabel>{t('mcp.toolCallTimeout')}</FormLabel>
-              <FormControl>
-                <Input
-                  type="number"
-                  min={0}
-                  step={1}
-                  placeholder="300"
-                  {...field}
-                  onChange={(e) => field.onChange(Number(e.target.value))}
-                />
-              </FormControl>
-              <FormDescription>
-                {t('mcp.toolCallTimeoutDescription')}
-              </FormDescription>
-              <FormMessage />
-            </FormItem>
-          )}
-        />
-
        {watchMode === 'remote' && (
          <>
            <FormField
@@ -1,18 +0,0 @@
-import {
-  AppWindow,
-  AudioWaveform,
-  Book,
-  FileText,
-  Hash,
-  Wrench,
-  type LucideIcon,
-} from 'lucide-react';
-
-export const pluginComponentIconMap: Record<string, LucideIcon> = {
-  Tool: Wrench,
-  EventListener: AudioWaveform,
-  Command: Hash,
-  KnowledgeEngine: Book,
-  Parser: FileText,
-  Page: AppWindow,
-};
@@ -1,7 +1,7 @@
 import { Fragment } from 'react';
 import { TFunction } from 'i18next';
+import { Wrench, AudioWaveform, Hash, Book, FileText } from 'lucide-react';
 import { Badge } from '@/components/ui/badge';
-import { pluginComponentIconMap } from './PluginComponentIcons';

 export default function PluginComponentList({
  components,
@@ -18,6 +18,14 @@ export default function PluginComponentList({
  t: TFunction;
  responsive?: boolean;
 }) {
+  const kindIconMap: Record<string, React.ReactNode> = {
+    Tool: <Wrench className="w-5 h-5" />,
+    EventListener: <AudioWaveform className="w-5 h-5" />,
+    Command: <Hash className="w-5 h-5" />,
+    KnowledgeEngine: <Book className="w-5 h-5" />,
+    Parser: <FileText className="w-5 h-5" />,
+  };
+
  const componentKindList = Object.keys(components || {});

  return (
@@ -26,12 +34,11 @@ export default function PluginComponentList({
      {componentKindList.length > 0 && (
        <>
          {componentKindList.map((kind) => {
-            const ComponentIcon = pluginComponentIconMap[kind];
            return (
              <Fragment key={kind}>
                {useBadge && (
                  <Badge variant="outline" className="flex items-center gap-1">
-                    {ComponentIcon && <ComponentIcon className="w-5 h-5" />}
+                    {kindIconMap[kind]}
                    {responsive ? (
                      <span className="hidden md:inline">
                        {t('market.componentName.' + kind)}
@@ -45,7 +52,7 @@ export default function PluginComponentList({

                {!useBadge && (
                  <div className="flex flex-row items-center justify-start gap-[0.2rem]">
-                    {ComponentIcon && <ComponentIcon className="w-5 h-5" />}
+                    {kindIconMap[kind]}
                    {responsive ? (
                      <span className="hidden md:inline">
                        {t('market.componentName.' + kind)}
@@ -127,7 +127,7 @@ function MarketPageContent({
  // Per-format extension counts shown next to the type filter options.
  const [typeCounts, setTypeCounts] = useState<Record<string, number>>({});
  const [sortOption, setSortOption] = useState<string>(
-    () => loadMarketFilters().sortOption ?? 'hot_score_desc',
+    () => loadMarketFilters().sortOption ?? 'install_count_desc',
  );

  // Persist filter conditions so they survive navigation / reload.
@@ -154,12 +154,6 @@ function MarketPageContent({

  // 排序选项
  const sortOptions: SortOption[] = [
-    {
-      value: 'hot_score_desc',
-      label: t('market.sort.hottest'),
-      sortBy: 'hot_score',
-      sortOrder: 'DESC',
-    },
    {
      value: 'created_at_desc',
      label: t('market.sort.recentlyAdded'),
@@ -217,7 +211,7 @@ function MarketPageContent({
    const option = sortOptions.find((opt) => opt.value === sortOption);
    return option
      ? { sortBy: option.sortBy, sortOrder: option.sortOrder }
-      : { sortBy: 'hot_score', sortOrder: 'DESC' };
+      : { sortBy: 'install_count', sortOrder: 'DESC' };
  }, [sortOption]);

  // 将API响应转换为VO对象
@@ -239,7 +233,6 @@ function MarketPageContent({
        description:
          extractI18nObject(plugin.description) || t('market.noDescription'),
        installCount: plugin.install_count || 0,
-        likeCount: plugin.like_count || 0,
        iconURL,
        githubURL: plugin.repository,
        version: plugin.latest_version,
@@ -40,7 +40,6 @@ function pluginToVO(
    description:
      extractI18nObject(plugin.description) || t('market.noDescription'),
    installCount: plugin.install_count,
-    likeCount: plugin.like_count || 0,
    iconURL,
    githubURL: plugin.repository,
    version: plugin.latest_version,
@@ -1,118 +0,0 @@
-import FingerprintJS, { type LoadOptions } from '@fingerprintjs/fingerprintjs';
-import { getCloudServiceClient } from '@/app/infra/http';
-
-const MARKETPLACE_LIKE_CHANGED_EVENT = 'langbot-marketplace-like-changed';
-
-export interface MarketplaceLikeChange {
-  key: string;
-  liked: boolean;
-  likeCount: number;
-}
-
-let fingerprintPromise: Promise<string> | undefined;
-let likedKeysPromise: Promise<Set<string>> | undefined;
-
-export function marketplaceExtensionKey(
-  type: string | undefined,
-  author: string,
-  name: string,
-): string {
-  return `${type || 'plugin'}:${author}/${name}`;
-}
-
-export function getMarketplaceFingerprint(): Promise<string> {
-  if (!fingerprintPromise) {
-    fingerprintPromise = FingerprintJS.load({
-      monitoring: false,
-    } as LoadOptions & { monitoring: boolean })
-      .then((agent) => agent.get())
-      .then((result) => result.visitorId)
-      .catch((error) => {
-        fingerprintPromise = undefined;
-        throw error;
-      });
-  }
-  return fingerprintPromise;
-}
-
-async function loadLikedKeys(): Promise<Set<string>> {
-  if (!likedKeysPromise) {
-    likedKeysPromise = Promise.all([
-      getMarketplaceFingerprint(),
-      getCloudServiceClient(),
-    ])
-      .then(([fingerprint, client]) =>
-        client.getMarketplaceLikedExtensions(fingerprint),
-      )
-      .then((data) => {
-        const keys = new Set<string>();
-        for (const ref of data.extensions || []) {
-          keys.add(`${ref.type}:${ref.extension_id}`);
-        }
-        return keys;
-      })
-      .catch((error) => {
-        likedKeysPromise = undefined;
-        throw error;
-      });
-  }
-  return likedKeysPromise;
-}
-
-export async function getMarketplaceExtensionLiked(
-  type: string | undefined,
-  author: string,
-  name: string,
-): Promise<boolean> {
-  const likedKeys = await loadLikedKeys();
-  return likedKeys.has(marketplaceExtensionKey(type, author, name));
-}
-
-export async function toggleMarketplaceExtensionLike(
-  type: string | undefined,
-  author: string,
-  name: string,
-  liked: boolean,
-): Promise<MarketplaceLikeChange> {
-  const extensionType = type || 'plugin';
-  const [fingerprint, likedKeys, client] = await Promise.all([
-    getMarketplaceFingerprint(),
-    loadLikedKeys(),
-    getCloudServiceClient(),
-  ]);
-  const result = await client.setMarketplaceExtensionLike(
-    extensionType,
-    author,
-    name,
-    fingerprint,
-    liked,
-  );
-  const key = marketplaceExtensionKey(extensionType, author, name);
-  if (result.liked) {
-    likedKeys.add(key);
-  } else {
-    likedKeys.delete(key);
-  }
-  const change: MarketplaceLikeChange = {
-    key,
-    liked: result.liked,
-    likeCount: result.like_count,
-  };
-  window.dispatchEvent(
-    new CustomEvent<MarketplaceLikeChange>(MARKETPLACE_LIKE_CHANGED_EVENT, {
-      detail: change,
-    }),
-  );
-  return change;
-}
-
-export function subscribeMarketplaceLikeChanges(
-  listener: (change: MarketplaceLikeChange) => void,
-): () => void {
-  const handler = (event: Event) => {
-    listener((event as CustomEvent<MarketplaceLikeChange>).detail);
-  };
-  window.addEventListener(MARKETPLACE_LIKE_CHANGED_EVENT, handler);
-  return () =>
-    window.removeEventListener(MARKETPLACE_LIKE_CHANGED_EVENT, handler);
-}
@@ -3,7 +3,7 @@ import { useRef, useState, useEffect } from 'react';
 import { useTranslation } from 'react-i18next';
 import PluginComponentList from '../PluginComponentList';
 import { Badge } from '@/components/ui/badge';
-import { Info, Package, ExternalLink, Heart, Loader2 } from 'lucide-react';
+import { Info, Package, ExternalLink } from 'lucide-react';
 import {
  Tooltip,
  TooltipContent,
@@ -11,13 +11,6 @@ import {
  TooltipTrigger,
 } from '@/components/ui/tooltip';
 import { Button } from '@/components/ui/button';
-import { toast } from 'sonner';
-import {
-  getMarketplaceExtensionLiked,
-  marketplaceExtensionKey,
-  subscribeMarketplaceLikeChanges,
-  toggleMarketplaceExtensionLike,
-} from '../marketplace-likes';

 export default function PluginMarketCardComponent({
  cardVO,
@@ -32,9 +25,6 @@ export default function PluginMarketCardComponent({
  const bottomRef = useRef<HTMLDivElement>(null);
  const [visibleTags, setVisibleTags] = useState(2);
  const [iconFailed, setIconFailed] = useState(!cardVO.iconURL);
-  const [liked, setLiked] = useState(false);
-  const [likeCount, setLikeCount] = useState(cardVO.likeCount);
-  const [isLiking, setIsLiking] = useState(false);

  const pluginDetailUrl = `https://space.langbot.app/market/${cardVO.author}/${cardVO.pluginName}`;

@@ -62,37 +52,6 @@ export default function PluginMarketCardComponent({
    setIconFailed(!cardVO.iconURL);
  }, [cardVO.iconURL]);

-  useEffect(() => {
-    setLikeCount(cardVO.likeCount);
-  }, [cardVO.likeCount]);
-
-  useEffect(() => {
-    let cancelled = false;
-    getMarketplaceExtensionLiked(cardVO.type, cardVO.author, cardVO.pluginName)
-      .then((value) => {
-        if (!cancelled) setLiked(value);
-      })
-      .catch(() => {
-        // The count still renders if the viewer-state request is unavailable.
-      });
-    return () => {
-      cancelled = true;
-    };
-  }, [cardVO.author, cardVO.pluginName, cardVO.type]);
-
-  useEffect(() => {
-    const key = marketplaceExtensionKey(
-      cardVO.type,
-      cardVO.author,
-      cardVO.pluginName,
-    );
-    return subscribeMarketplaceLikeChanges((change) => {
-      if (change.key !== key) return;
-      setLiked(change.liked);
-      setLikeCount(change.likeCount);
-    });
-  }, [cardVO.author, cardVO.pluginName, cardVO.type]);
-
  useEffect(() => {
    const tags = cardVO.tags;
    if (!bottomRef.current || !tags || tags.length === 0) return;
@@ -130,29 +89,6 @@ export default function PluginMarketCardComponent({
    onInstall?.(cardVO);
  };

-  const handleLikeClick = async (
-    event: React.MouseEvent<HTMLButtonElement>,
-  ) => {
-    event.preventDefault();
-    event.stopPropagation();
-    if (isLiking) return;
-    setIsLiking(true);
-    try {
-      const change = await toggleMarketplaceExtensionLike(
-        cardVO.type,
-        cardVO.author,
-        cardVO.pluginName,
-        !liked,
-      );
-      setLiked(change.liked);
-      setLikeCount(change.likeCount);
-    } catch {
-      toast.error(t('market.likeFailed'));
-    } finally {
-      setIsLiking(false);
-    }
-  };
-
  return (
    <div
      role="button"
@@ -161,10 +97,7 @@ export default function PluginMarketCardComponent({
      className="w-[100%] h-[10rem] cursor-pointer bg-white rounded-[10px] border border-border shadow-[0px_1px_2px_0_rgba(0,0,0,0.06)] p-3 sm:p-[1rem] hover:shadow-[0px_2px_5px_0_rgba(0,0,0,0.08)] transition-shadow duration-200 outline-none focus-visible:ring-[3px] focus-visible:ring-ring/50 dark:bg-[#1f1f22] dark:shadow-[0px_1px_2px_0_rgba(255,255,255,0.04)] dark:hover:shadow-[0px_2px_5px_0_rgba(255,255,255,0.07)] relative"
      onClick={handleInstallClick}
      onKeyDown={(event) => {
-        if (
-          event.target === event.currentTarget &&
-          (event.key === 'Enter' || event.key === ' ')
-        ) {
+        if (event.key === 'Enter' || event.key === ' ') {
          event.preventDefault();
          handleInstallClick();
        }
@@ -244,26 +177,6 @@ export default function PluginMarketCardComponent({
          </div>

          <div className="flex flex-row items-start justify-center gap-1 flex-shrink-0">
-            <Button
-              type="button"
-              variant="ghost"
-              size="sm"
-              title={t(liked ? 'market.unlike' : 'market.like')}
-              aria-label={t(liked ? 'market.unlike' : 'market.like')}
-              aria-pressed={liked}
-              disabled={isLiking}
-              className="h-7 min-w-7 gap-1 rounded-md px-1.5 text-muted-foreground hover:bg-red-50 hover:text-red-500 dark:hover:bg-red-950/30"
-              onClick={handleLikeClick}
-            >
-              {isLiking ? (
-                <Loader2 className="h-3.5 w-3.5 animate-spin" />
-              ) : (
-                <Heart
-                  className={`h-3.5 w-3.5 ${liked ? 'fill-red-500 text-red-500' : ''}`}
-                />
-              )}
-              <span className="text-xs tabular-nums">{likeCount}</span>
-            </Button>
            <Button
              type="button"
              variant="ghost"
@@ -5,7 +5,6 @@ export interface IPluginMarketCardVO {
  label: string;
  description: string;
  installCount: number;
-  likeCount?: number;
  iconURL: string;
  githubURL: string;
  version: string;
@@ -23,7 +22,6 @@ export class PluginMarketCardVO implements IPluginMarketCardVO {
  iconURL: string;
  githubURL: string;
  installCount: number;
-  likeCount: number;
  version: string;
  components?: Record<string, number>;
  tags?: string[];
@@ -37,7 +35,6 @@ export class PluginMarketCardVO implements IPluginMarketCardVO {
    this.iconURL = prop.iconURL;
    this.githubURL = prop.githubURL;
    this.installCount = prop.installCount;
-    this.likeCount = prop.likeCount ?? 0;
    this.pluginId = prop.pluginId;
    this.version = prop.version;
    this.components = prop.components;
@@ -519,21 +519,18 @@ export interface MCPServerExtraArgsSSE {
  headers: Record<string, string>;
  timeout: number;
  ssereadtimeout: number;
-  tool_call_timeout_sec?: number;
 }

 export interface MCPServerExtraArgsStdio {
  command: string;
  args: string[];
  env: Record<string, string>;
-  tool_call_timeout_sec?: number;
 }

 export interface MCPServerExtraArgsHttp {
  url: string;
  headers: Record<string, string>;
  timeout: number;
-  tool_call_timeout_sec?: number;
 }

 // "remote" mode: the user only supplies a URL; the backend auto-detects the
@@ -543,7 +540,6 @@ export interface MCPServerExtraArgsRemote {
  url: string;
  headers?: Record<string, string>;
  timeout?: number;
-  tool_call_timeout_sec?: number;
 }

 export enum MCPSessionStatus {
@@ -48,8 +48,6 @@ export interface PluginV4 {
  repository: string;
  tags: string[];
  install_count: number;
-  like_count?: number;
-  hot_score?: number;
  latest_version: string;
  components: Record<string, number>;
  status: PluginV4Status;
@@ -117,32 +117,6 @@ export class CloudServiceClient extends BaseHttpClient {
      .catch(() => this.searchMarketplaceExtensionsLegacy(data));
  }

-  public getMarketplaceLikedExtensions(
-    fingerprint: string,
-  ): Promise<{ extensions: Array<{ type: string; extension_id: string }> }> {
-    return this.post('/api/v1/marketplace/extensions/likes/status', {
-      fingerprint,
-    });
-  }
-
-  public setMarketplaceExtensionLike(
-    type: string,
-    author: string,
-    name: string,
-    fingerprint: string,
-    liked: boolean,
-  ): Promise<{
-    type: string;
-    extension_id: string;
-    liked: boolean;
-    like_count: number;
-  }> {
-    return this.put(
-      `/api/v1/marketplace/extensions/${encodeURIComponent(type)}/${encodeURIComponent(author)}/${encodeURIComponent(name)}/like`,
-      { fingerprint, liked },
-    );
-  }
-
  private async searchMarketplaceExtensionsLegacy(data: {
    query?: string;
    page: number;
@@ -154,8 +128,6 @@ export class CloudServiceClient extends BaseHttpClient {
    tags_filter?: string[];
  }): Promise<ApiRespMarketplacePlugins> {
    const query = data.query || '';
-    const legacySortBy =
-      data.sort_by === 'hot_score' ? 'install_count' : data.sort_by;

    if (
      data.type_filter === 'plugin' ||
@@ -167,7 +139,7 @@ export class CloudServiceClient extends BaseHttpClient {
        query,
        data.page,
        data.page_size,
-        legacySortBy,
+        data.sort_by,
        data.sort_order,
        data.component_filter,
        data.tags_filter,
@@ -185,7 +157,7 @@ export class CloudServiceClient extends BaseHttpClient {
        query,
        data.page,
        data.page_size,
-        legacySortBy,
+        data.sort_by,
        data.sort_order,
        undefined,
        data.tags_filter,
@@ -195,7 +167,7 @@ export class CloudServiceClient extends BaseHttpClient {
        query,
        data.page,
        data.page_size,
-        legacySortBy,
+        data.sort_by,
        data.sort_order,
        undefined,
        data.tags_filter,
@@ -205,7 +177,7 @@ export class CloudServiceClient extends BaseHttpClient {
        query,
        data.page,
        data.page_size,
-        legacySortBy,
+        data.sort_by,
        data.sort_order,
        undefined,
        data.tags_filter,
@@ -700,7 +700,6 @@ const enUS = {
    notFound: 'Plugin information not found',
    sortBy: 'Sort by',
    sort: {
-      hottest: 'Most Popular',
      recentlyAdded: 'Recently Added',
      recentlyUpdated: 'Recently Updated',
      mostDownloads: 'Most Downloads',
@@ -708,9 +707,6 @@ const enUS = {
    },
    downloads: 'downloads',
    download: 'Download',
-    like: 'Like',
-    unlike: 'Unlike',
-    likeFailed: 'Failed to update like. Please try again.',
    repository: 'Repository',
    downloadFailed: 'Download failed',
    noReadme: 'This plugin does not provide README documentation',
@@ -801,9 +797,6 @@ const enUS = {
    url: 'URL',
    headers: 'Headers',
    timeout: 'Timeout',
-    toolCallTimeout: 'Tool call timeout (seconds)',
-    toolCallTimeoutDescription:
-      'Maximum wait for one tool call. Set to 0 for no timeout. Defaults to 300 seconds.',
    addArgument: 'Add Argument',
    addEnvVar: 'Add Environment Variable',
    addHeader: 'Add Header',
@@ -713,7 +713,6 @@ const esES = {
    notFound: 'No se encontró la información del plugin',
    sortBy: 'Ordenar por',
    sort: {
-      hottest: 'Más populares',
      recentlyAdded: 'Añadidos recientemente',
      recentlyUpdated: 'Actualizados recientemente',
      mostDownloads: 'Más descargas',
@@ -721,9 +720,6 @@ const esES = {
    },
    downloads: 'descargas',
    download: 'Descargar',
-    like: 'Me gusta',
-    unlike: 'Ya no me gusta',
-    likeFailed: 'No se pudo actualizar el Me gusta. Inténtalo de nuevo.',
    repository: 'Repositorio',
    downloadFailed: 'Error en la descarga',
    noReadme: 'Este plugin no proporciona documentación README',
@@ -815,9 +811,6 @@ const esES = {
    url: 'URL',
    headers: 'Encabezados',
    timeout: 'Tiempo de espera',
-    toolCallTimeout: 'Tiempo de espera de herramienta (segundos)',
-    toolCallTimeoutDescription:
-      'Espera máxima para una llamada de herramienta. Use 0 para no limitar. El valor predeterminado es 300 segundos.',
    addArgument: 'Añadir argumento',
    addEnvVar: 'Añadir variable de entorno',
    addHeader: 'Añadir encabezado',
@@ -706,7 +706,6 @@ const jaJP = {
    notFound: 'プラグイン情報が見つかりません',
    sortBy: '並び順',
    sort: {
-      hottest: '人気順',
      recentlyAdded: '最近追加',
      recentlyUpdated: '最近更新',
      mostDownloads: 'ダウンロード数多',
@@ -714,9 +713,6 @@ const jaJP = {
    },
    downloads: '回ダウンロード',
    download: 'ダウンロード',
-    like: 'いいね',
-    unlike: 'いいねを解除',
-    likeFailed: 'いいねを更新できませんでした。もう一度お試しください。',
    repository: 'リポジトリ',
    downloadFailed: 'ダウンロード失敗',
    noReadme: 'このプラグインはREADMEドキュメントを提供していません',
@@ -807,9 +803,6 @@ const jaJP = {
    url: 'URL',
    headers: 'ヘッダー',
    timeout: 'タイムアウト',
-    toolCallTimeout: 'ツール呼び出しタイムアウト（秒）',
-    toolCallTimeoutDescription:
-      '1 回のツール呼び出しの最大待機時間です。0 で無制限、既定値は 300 秒です。',
    addArgument: '引数を追加',
    addEnvVar: '環境変数を追加',
    addHeader: 'ヘッダーを追加',
@@ -711,7 +711,6 @@ const ruRU = {
    notFound: 'Информация о плагине не найдена',
    sortBy: 'Сортировать по',
    sort: {
-      hottest: 'По популярности',
      recentlyAdded: 'Недавно добавленные',
      recentlyUpdated: 'Недавно обновлённые',
      mostDownloads: 'Больше всего загрузок',
@@ -719,9 +718,6 @@ const ruRU = {
    },
    downloads: 'загрузок',
    download: 'Скачать',
-    like: 'Нравится',
-    unlike: 'Убрать отметку',
-    likeFailed: 'Не удалось обновить отметку. Повторите попытку.',
    repository: 'Репозиторий',
    downloadFailed: 'Ошибка загрузки',
    noReadme: 'Этот плагин не предоставляет документацию README',
@@ -812,9 +808,6 @@ const ruRU = {
    url: 'URL',
    headers: 'Заголовки',
    timeout: 'Таймаут',
-    toolCallTimeout: 'Таймаут вызова инструмента (секунды)',
-    toolCallTimeoutDescription:
-      'Максимальное ожидание одного вызова. 0 отключает ограничение. По умолчанию 300 секунд.',
    addArgument: 'Добавить аргумент',
    addEnvVar: 'Добавить переменную окружения',
    addHeader: 'Добавить заголовок',
@@ -690,7 +690,6 @@ const thTH = {
    notFound: 'ไม่พบข้อมูลปลั๊กอิน',
    sortBy: 'เรียงตาม',
    sort: {
-      hottest: 'ยอดนิยมที่สุด',
      recentlyAdded: 'เพิ่มล่าสุด',
      recentlyUpdated: 'อัปเดตล่าสุด',
      mostDownloads: 'ดาวน์โหลดมากที่สุด',
@@ -698,9 +697,6 @@ const thTH = {
    },
    downloads: 'ดาวน์โหลด',
    download: 'ดาวน์โหลด',
-    like: 'ถูกใจ',
-    unlike: 'เลิกถูกใจ',
-    likeFailed: 'อัปเดตการถูกใจไม่สำเร็จ โปรดลองอีกครั้ง',
    repository: 'Repository',
    downloadFailed: 'ดาวน์โหลดล้มเหลว',
    noReadme: 'ปลั๊กอินนี้ไม่มีเอกสาร README',
@@ -790,9 +786,6 @@ const thTH = {
    url: 'URL',
    headers: 'ส่วนหัว',
    timeout: 'หมดเวลา',
-    toolCallTimeout: 'หมดเวลาการเรียกเครื่องมือ (วินาที)',
-    toolCallTimeoutDescription:
-      'เวลารอสูงสุดต่อการเรียกเครื่องมือหนึ่งครั้ง ตั้งเป็น 0 เพื่อไม่จำกัด ค่าเริ่มต้นคือ 300 วินาที',
    addArgument: 'เพิ่มอาร์กิวเมนต์',
    addEnvVar: 'เพิ่มตัวแปรสภาพแวดล้อม',
    addHeader: 'เพิ่มส่วนหัว',
@@ -705,7 +705,6 @@ const viVN = {
    notFound: 'Không tìm thấy thông tin plugin',
    sortBy: 'Sắp xếp theo',
    sort: {
-      hottest: 'Phổ biến nhất',
      recentlyAdded: 'Mới thêm gần đây',
      recentlyUpdated: 'Mới cập nhật gần đây',
      mostDownloads: 'Tải nhiều nhất',
@@ -713,9 +712,6 @@ const viVN = {
    },
    downloads: 'lượt tải',
    download: 'Tải xuống',
-    like: 'Thích',
-    unlike: 'Bỏ thích',
-    likeFailed: 'Không thể cập nhật lượt thích. Vui lòng thử lại.',
    repository: 'Kho lưu trữ',
    downloadFailed: 'Tải xuống thất bại',
    noReadme: 'Plugin này không cung cấp tài liệu README',
@@ -805,9 +801,6 @@ const viVN = {
    url: 'URL',
    headers: 'Tiêu đề',
    timeout: 'Thời gian chờ',
-    toolCallTimeout: 'Thời gian chờ gọi công cụ (giây)',
-    toolCallTimeoutDescription:
-      'Thời gian chờ tối đa cho một lần gọi công cụ. Đặt 0 để không giới hạn. Mặc định là 300 giây.',
    addArgument: 'Thêm tham số',
    addEnvVar: 'Thêm biến môi trường',
    addHeader: 'Thêm tiêu đề',
@@ -668,7 +668,6 @@ const zhHans = {
    notFound: '插件信息未找到',
    sortBy: '排序方式',
    sort: {
-      hottest: '热度最高',
      recentlyAdded: '最近新增',
      recentlyUpdated: '最近更新',
      mostDownloads: '最多下载',
@@ -676,9 +675,6 @@ const zhHans = {
    },
    downloads: '次下载',
    download: '下载',
-    like: '点赞',
-    unlike: '取消点赞',
-    likeFailed: '点赞失败，请稍后重试',
    repository: '代码仓库',
    downloadFailed: '下载失败',
    noReadme: '该插件没有提供 README 文档',
@@ -767,9 +763,6 @@ const zhHans = {
    url: 'URL地址',
    headers: '请求头',
    timeout: '超时时间',
-    toolCallTimeout: '工具调用超时（秒）',
-    toolCallTimeoutDescription:
-      '单次工具调用的最长等待时间。设为 0 表示不限制，默认 300 秒。',
    addArgument: '添加参数',
    addEnvVar: '添加环境变量',
    addHeader: '添加请求头',
@@ -668,7 +668,6 @@ const zhHant = {
    notFound: '插件資訊未找到',
    sortBy: '排序方式',
    sort: {
-      hottest: '熱度最高',
      recentlyAdded: '最近新增',
      recentlyUpdated: '最近更新',
      mostDownloads: '最多下載',
@@ -676,9 +675,6 @@ const zhHant = {
    },
    downloads: '次下載',
    download: '下載',
-    like: '按讚',
-    unlike: '取消按讚',
-    likeFailed: '按讚失敗，請稍後重試',
    repository: '代碼倉庫',
    downloadFailed: '下載失敗',
    noReadme: '該插件沒有提供 README 文件',
@@ -766,9 +762,6 @@ const zhHant = {
    url: 'URL位址',
    headers: '請求標頭',
    timeout: '逾時時間',
-    toolCallTimeout: '工具呼叫逾時（秒）',
-    toolCallTimeoutDescription:
-      '單次工具呼叫的最長等待時間。設為 0 表示不限制，預設 300 秒。',
    addArgument: '新增參數',
    addEnvVar: '新增環境變數',
    addHeader: '新增請求標頭',
@@ -1,51 +0,0 @@
-import assert from 'node:assert/strict';
-import fs from 'node:fs';
-import path from 'node:path';
-import test from 'node:test';
-import ts from 'typescript';
-import { fileURLToPath } from 'node:url';
-
-const currentDirectory = path.dirname(fileURLToPath(import.meta.url));
-const sourcePath = path.resolve(
-  currentDirectory,
-  '../../src/app/home/components/dynamic-form/DynamicFormSaveValues.ts',
-);
-
-function loadNormalizer() {
-  const source = fs.readFileSync(sourcePath, 'utf8');
-  const compiled = ts.transpileModule(source, {
-    compilerOptions: { module: ts.ModuleKind.CommonJS },
-  }).outputText;
-  const loadedModule = { exports: {} };
-  new Function('require', 'module', 'exports', compiled)(
-    () => {
-      throw new Error('DynamicFormSaveValues must not have runtime imports');
-    },
-    loadedModule,
-    loadedModule.exports,
-  );
-  return loadedModule.exports.normalizeDynamicFormValuesForSave;
-}
-
-test('normalizes only single-line text fields in a dynamic form save snapshot', () => {
-  const normalizeDynamicFormValuesForSave = loadNormalizer();
-  const specs = [
-    { name: 'single-line', type: 'string', default: '' },
-    { name: 'multiline', type: 'text', default: '' },
-    { name: 'string-list', type: 'array[string]', default: [] },
-    { name: 'count', type: 'integer', default: 0 },
-  ];
-  const values = {
-    'single-line': '\t  hello  world \n',
-    multiline: '  keep multiline whitespace  \n',
-    'string-list': ['  first  ', ' second '],
-    count: 3,
-  };
-
-  assert.deepEqual(normalizeDynamicFormValuesForSave(specs, values), {
-    'single-line': 'hello  world',
-    multiline: '  keep multiline whitespace  \n',
-    'string-list': ['  first  ', ' second '],
-    count: 3,
-  });
-});
@@ -1,86 +0,0 @@
-import assert from 'node:assert/strict';
-import fs from 'node:fs';
-import path from 'node:path';
-import test from 'node:test';
-import { fileURLToPath } from 'node:url';
-
-const webRoot = path.resolve(
-  path.dirname(fileURLToPath(import.meta.url)),
-  '../..',
-);
-const read = (relativePath) =>
-  fs.readFileSync(path.join(webRoot, relativePath), 'utf8');
-
-test('marketplace defaults to shared hot sorting and cards support fingerprint likes', () => {
-  const market = read(
-    'src/app/home/plugins/components/plugin-market/PluginMarketComponent.tsx',
-  );
-  const card = read(
-    'src/app/home/plugins/components/plugin-market/plugin-market-card/PluginMarketCardComponent.tsx',
-  );
-  const cardVO = read(
-    'src/app/home/plugins/components/plugin-market/plugin-market-card/PluginMarketCardVO.ts',
-  );
-  const entity = read('src/app/infra/entities/plugin/index.ts');
-  const client = read('src/app/infra/http/CloudServiceClient.ts');
-  const likes = read(
-    'src/app/home/plugins/components/plugin-market/marketplace-likes.ts',
-  );
-
-  assert.match(
-    market,
-    /hot_score_desc/,
-    'market must expose and default to hot-score sorting',
-  );
-  assert.match(
-    market,
-    /sortBy:\s*['"]hot_score['"]/,
-    'hot option must use the shared API field',
-  );
-  assert.match(
-    entity,
-    /like_count\??:\s*number/,
-    'marketplace extension responses must carry like counts',
-  );
-  assert.match(
-    cardVO,
-    /likeCount:\s*number/,
-    'market cards must carry like counts',
-  );
-  assert.match(card, /\bHeart\b/, 'market cards must render a like affordance');
-  assert.match(
-    card,
-    /toggleMarketplaceExtensionLike/,
-    'market cards must support liking and unliking',
-  );
-  assert.match(
-    client,
-    /marketplace\/extensions\/likes/,
-    'client must load browser likes',
-  );
-  assert.match(
-    client,
-    /setMarketplaceExtensionLike\(/,
-    'client must update likes through the shared API',
-  );
-  assert.match(
-    client,
-    /data\.sort_by === ['"]hot_score['"] \? ['"]install_count['"] : data\.sort_by/,
-    'older Space servers must fall back from hot score to install sorting',
-  );
-  assert.match(
-    likes,
-    /@fingerprintjs\/fingerprintjs/,
-    'anonymous likes must use browser fingerprinting',
-  );
-  assert.match(
-    likes,
-    /FingerprintJS\.load\(\{/,
-    'fingerprint agent must be initialized with options',
-  );
-  assert.match(
-    likes,
-    /monitoring:\s*false/,
-    'fingerprinting must not send vendor monitoring requests',
-  );
-});
@@ -1,31 +0,0 @@
-import assert from 'node:assert/strict';
-import fs from 'node:fs';
-import path from 'node:path';
-import test from 'node:test';
-import { createRequire } from 'node:module';
-import { fileURLToPath } from 'node:url';
-import ts from 'typescript';
-
-const currentDirectory = path.dirname(fileURLToPath(import.meta.url));
-const sourcePath = path.resolve(
-  currentDirectory,
-  '../../src/app/home/plugins/components/plugin-market/PluginComponentIcons.ts',
-);
-const source = fs.readFileSync(sourcePath, 'utf8');
-const compiled = ts.transpileModule(source, {
-  compilerOptions: { module: ts.ModuleKind.CommonJS },
-}).outputText;
-const sourceRequire = createRequire(sourcePath);
-const loadedModule = { exports: {} };
-new Function('require', 'module', 'exports', compiled)(
-  sourceRequire,
-  loadedModule,
-  loadedModule.exports,
-);
-
-const { AppWindow } = sourceRequire('lucide-react');
-const { pluginComponentIconMap } = loadedModule.exports;
-
-test('maps the Page plugin component to the AppWindow icon', () => {
-  assert.equal(pluginComponentIconMap.Page, AppWindow);
-});
Author	SHA1	Message	Date
fdc310	1bdf371b5d	feat: Add pipeline_uuid handling across multiple adapters and update related tests	2026-07-12 19:40:59 +08:00
fdc310	cdfde83e39	feat: Add missing completed input lines handling in DingTalk adapter	2026-07-12 18:53:13 +08:00
fdc310	4a753d98f5	Merge remote-tracking branch 'origin/master' into feat/card_dify_human_input	2026-07-12 18:39:48 +08:00
fdc310	354aadfce0	feat: Enhance select input handling and validation in Dify API runner and Telegram adapter	2026-07-12 00:31:12 +08:00
fdc310	4b6931d294	feat: Enhance error handling and caching for Dify and WeCom interactions	2026-07-11 22:40:08 +08:00
fdc310	5eb2b90ae6	feat: Refactor pending form handling to isolate by bot and pipeline	2026-07-11 20:46:42 +08:00
fdc310	d8c14fc969	feat: Update the human-computer interaction template cards, and optimize the prompt information and content display.	2026-07-11 20:05:05 +08:00
fdc310	6cae954a67	feat: Refactor input handling and feedback messages across multiple adapters	2026-07-11 19:23:42 +08:00
fdc310	183fb87649	Add tests for DingTalk adapter content processing and markdown formatting - Updated the assertion in `test_dingtalk_completed_input_lines_include_text_and_select_values` to remove unnecessary markdown formatting. - Added new tests to verify that `_dingtalk_clean_form_content` maintains the order of prompts and completed values in various scenarios. - Introduced `test_dingtalk_card_markdown_preserves_internal_line_breaks` to ensure internal line breaks are correctly converted to HTML line breaks.	2026-07-11 17:28:31 +08:00
fdc310	56b3c15213	feat(lark): add functions for current input definitions and visible form content handling feat(qqofficial): update fallback text handling for non-streaming scenarios feat(difysvapi): enhance form content processing for interactive fields and actions test: add unit tests for Lark and QQ Official adapter functionalities	2026-07-11 15:31:29 +08:00
fdc310	0cde4c71cd	feat: Enhance Telegram and QQ Official adapters with select field handling and form action processing - Added support for select fields in Telegram adapter, including option extraction and callback handling. - Implemented form action processing for Telegram callbacks, improving user interaction feedback. - Introduced new helper functions for building keyboards and resolving select button actions in QQ Official adapter. - Enhanced DifyServiceAPIRunner to handle cumulative streaming responses and improve error handling during workflow resumes. - Added unit tests for new functionalities in Telegram and QQ Official adapters, ensuring robust behavior for select fields and form actions.	2026-07-11 11:16:59 +08:00
fdc310	98a633c712	Add unit tests for DingTalk, Lark, WeComBot, and Dify service API runners - Implement tests for DingTalk adapter helper functions including form content cleaning, input extraction, and completed input lines. - Create unit tests for Lark adapter helper functions focusing on input extraction and completed input lines. - Add tests for WeComBot template card functionalities, including event extraction and payload building for human input. - Enhance Dify service API runner tests to cover human input forms, including input collection, action handling, and form snapshot extraction.	2026-07-10 18:25:24 +08:00
fdc310	776d48380e	feat(telegram): enhance message handling with group stream deletion and form placeholder detection	2026-06-24 14:14:14 +08:00
fdc310	5e6b3cc550	feat(dingtalk): add download link for human input card template and enhance dynamic form configuration	2026-06-24 11:23:02 +08:00
fdc310	ea4cd80939	feat(qqofficial): add method to identify invisible form placeholder chunks in messages	2026-06-23 16:34:38 +08:00
fdc310	ff7fd2f5e0	feat(qqofficial): synchronize passive-reply counter with inbound message sequence	2026-06-23 16:04:29 +08:00
fdc310	3b3aa93cf8	feat(wecom): add functions for template card action extraction and update, enhance button interaction handling	2026-06-23 15:25:39 +08:00
fdc310	f658f0ede5	feat(wecom): add optional source block to interactive template cards for enhanced branding	2026-06-23 00:02:37 +08:00
fdc310	47d75082e8	fix: add Avatar component to dingtalk_human_input_card.json for enhanced user interaction	2026-06-22 20:16:10 +08:00
fdc310	8849fb172e	feat(dingtalk): add feedback functionality to cards with 👍/👎 buttons - Implemented feedback state management for cards, allowing users to provide feedback via thumbs up/down buttons. - Enhanced card rendering to include feedback buttons when appropriate. - Registered feedback listeners to handle feedback events and update card states accordingly. - Updated the card template to support dynamic button rendering for feedback actions. - Improved error handling and logging for feedback actions and card updates.	2026-06-22 15:48:35 +08:00
fdc310	6818ae2783	feat(dingtalk): update display formatting for card content with HTML line breaks	2026-06-21 17:57:39 +08:00
fdc310	e7f1059194	feat(lark): enhance LarkAdapter to store form content for resume notices	2026-06-21 17:17:59 +08:00
fdc310	537677044e	feat(telegram): add action title feedback for user selections in Telegram messages	2026-06-21 15:12:47 +08:00
fdc310	321891441c	feat(dify): add support for chatflow app type and enhance human input handling	2026-06-21 13:50:50 +08:00
fdc310	7f36103aa4	fix(telegram): correct group chat type check and handle oversized callback data for Telegram actions fix(difysvapi): ensure safe access to remove-think configuration in pipeline settings	2026-06-18 11:53:50 +08:00
fdc310	b1750882ed	feat(discord): implement Discord form view with button interactions for Dify actions	2026-06-17 11:25:59 +08:00
fdc310	b55f073e62	feat(qqofficial): implement one-click QR binding and enhance localization support	2026-06-17 00:31:29 +08:00
fdc310	e6cfee541f	feat(qqofficial): implement Dify human input button interaction handling and markdown keyboard support	2026-06-16 19:37:16 +08:00
fdc310	f10ea82418	feat(wecom): implement Dify human input pause handling with button interaction support	2026-06-16 00:07:11 +08:00
fdc310	a32c4d152f	feat(dingtalk): enhance human input card functionality with streaming support and active turn management - Updated the DingTalk card template to enable streaming mode and multi-update configuration. - Removed the obsolete delete_card method from DingTalkClient to streamline card management. - Enhanced DingTalkAdapter to manage active turn cards and accumulated streaming text, ensuring a seamless user experience during human input prompts. - Modified the create_message_card method to utilize existing active cards for resumed workflows, preventing duplication. - Improved the _paint_form_on_card method to update existing cards with human input prompts and buttons dynamically. - Updated the dingtalk_human_input_card.json template to reflect the new streaming capabilities and configuration options.	2026-06-15 17:45:09 +08:00
fdc310	83b0d26e99	feat(dingtalk): implement human input card support and card action handling - Add a new module `card_callback.py` to handle card action button clicks from DingTalk. - Introduce `DingTalkCardActionHandler` to process card action callbacks and extract parameters. - Update `DingTalkAdapter` to manage card state and handle form input through a single card template. - Add configuration for `human_input_card_template_id` in `dingtalk.yaml` to specify the template for human input. - Create a new card template `dingtalk_human_input_card.json` for rendering human input prompts and buttons.	2026-06-15 15:51:52 +08:00
fdc310	e08b5db625	feat: Add the function for formatting human input text to support adapters without rich UI.	2026-06-10 15:41:18 +08:00
fdc310	d0f65b17ec	feat: Improve TelegramAdapter message handling with enhanced error management and draft message support	2026-06-02 01:14:25 +08:00
fdc310	2b533c4a00	feat: Enhance TelegramAdapter to handle form action buttons and message threading	2026-06-02 00:25:52 +08:00
fdc310	f663d87a60	feat: Enhance Lark and Telegram adapters with new form handling for paused workflows	2026-06-01 23:48:59 +08:00
fdc310	60e5b873ee	feat: Add '_routed_by_rule' variable to form action in Lark and Telegram adapters	2026-05-30 12:22:00 +08:00
fdc310	b96f209b98	feat: Implement workflow form handling for paused workflows - Added module-level storage for pending forms to manage state across sessions. - Introduced functions to set, get, and clear pending forms with expiration handling. - Enhanced DifyServiceAPIRunner to support resuming paused workflows via form actions. - Implemented logic to yield human input requests and display appropriate messages. - Updated workflow submission methods to handle paused states and resume actions. - Ensured proper merging of pending form actions with user inputs for seamless interaction.	2026-05-28 23:32:46 +08:00