id: local-agent-basic-debug-chat
title: "Local Agent Debug Chat returns a deterministic streaming response"
mode: agent-browser
area: pipeline
type: smoke
priority: p0
risk: medium
ci_eligible: false
tags:
  - local-agent
  - pipeline
  - streaming
skills:
  - langbot-env-setup
  - langbot-testing
env:
  - LANGBOT_FRONTEND_URL
  - LANGBOT_BACKEND_URL
  - LANGBOT_LOCAL_AGENT_PIPELINE_URL
  - LANGBOT_LOCAL_AGENT_PIPELINE_NAME
automation: scripts/e2e/pipeline-debug-chat.mjs
automation_env:
  - LANGBOT_FRONTEND_URL
  - LANGBOT_BROWSER_PROFILE
  - LANGBOT_CHROMIUM_EXECUTABLE
  - LANGBOT_LOCAL_AGENT_PIPELINE_URL
  - LANGBOT_LOCAL_AGENT_PIPELINE_NAME
automation_pipeline_url_env: LANGBOT_LOCAL_AGENT_PIPELINE_URL
automation_pipeline_name_env: LANGBOT_LOCAL_AGENT_PIPELINE_NAME
automation_prompt: "请只回复 OK，用于前端调试测试。"
automation_expected_text: "OK"
setup_automation:
  - "node:scripts/e2e/ensure-local-agent-pipeline.mjs --write-env"
setup_provides_env:
  - LANGBOT_LOCAL_AGENT_PIPELINE_URL
  - LANGBOT_LOCAL_AGENT_PIPELINE_NAME
steps:
  - "Open LANGBOT_FRONTEND_URL."
  - "Navigate to Pipelines and open the target local-agent pipeline."
  - "Open Configuration > AI."
  - "Use runner Default or the pluginized langbot-team/LocalAgent runner."
  - "Select a model that is known to answer Debug Chat in the current environment."
  - "Clear Knowledge Bases unless this run intentionally combines RAG with the smoke path."
  - "Save the pipeline."
  - "Open Debug Chat."
  - "Ensure the stream switch is enabled when the UI exposes it."
  - "Send: 请只回复 OK，用于前端调试测试。"
checks:
  - "UI: The user message appears in Debug Chat."
  - "UI: A Bot message appears and contains OK."
  - "Console: No unexpected frontend runtime errors appear during the send/receive path."
  - "Logs: Backend logs show the debug chat request completed on the streaming path instead of timing out in plugin/runtime calls."
evidence_required:
  - ui
  - screenshot
  - console
  - backend_log
diagnostics:
  - "Provider errors such as model_not_found or no available channel mean the selected model is unavailable; switch to a known-good model before diagnosing local-agent."
  - "Use GET /api/v1/pipelines/{uuid} only to confirm the saved runner and model config."
success_patterns:
  - "Processing request from person_websocket"
  - "Streaming completed"
failure_patterns:
  - "Action invoke_llm_stream call timed out"
  - "Task exception was never retrieved"
  - "survey widget blocks debug chat"
troubleshooting:
  - plugin-runtime-timeout
  - proxy-env-mismatch
  - survey-widget-blocks-debug-chat