chore(release): bump version to 4.10.2

Load the instance creation timestamp from data/labels/instance_id.json
(backfilling+persisting it for instances created before the field existed),
expose it as constants.instance_create_ts, and include it in the heartbeat
payload so Space can anchor Time-To-Value / onboarding analytics on real
install time rather than first-heartbeat.

Verified: py_compile, ruff, pytest tests/unit_tests/telemetry/ (37 passed).

pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "langbot"
-version = "4.10.1"
+version = "4.10.2"
 description = "Production-grade platform for building agentic IM bots"
 readme = "README.md"
 license-files = ["LICENSE"]

src/langbot/__init__.py

@@ -1,3 +1,3 @@
 """LangBot - Production-grade platform for building agentic IM bots"""
 
-__version__ = '4.10.1'
+__version__ = '4.10.2'

src/langbot/pkg/core/stages/load_config.py

@@ -202,6 +202,16 @@ class LoadConfigStage(stage.BootingStage):
                 constants.instance_id = new_id
         constants.edition = ap.instance_config.data.get('system', {}).get('edition', 'community')
 
+        # Instance creation timestamp: sourced from data/labels/instance_id.json.
+        # Instances created before this field existed (or supplied via
+        # system.instance_id) won't have it, so backfill with the current time
+        # and persist it via the dump below — from then on it stays stable.
+        instance_create_ts = ap.instance_id.data.get('instance_create_ts', 0)
+        if not isinstance(instance_create_ts, int) or instance_create_ts <= 0:
+            instance_create_ts = int(time.time())
+            ap.instance_id.data['instance_create_ts'] = instance_create_ts
+        constants.instance_create_ts = instance_create_ts
+
         print(f'LangBot instance id: {constants.instance_id}')
         print(f'LangBot edition: {constants.edition}')
 

src/langbot/pkg/provider/modelmgr/requesters/litellmchat.py

@@ -75,22 +75,33 @@ class LiteLLMRequester(requester.ProviderAPIRequester):
                 continue
         return False
 
+    @staticmethod
+    def _positive_int(value: typing.Any) -> int | None:
+        if isinstance(value, bool):
+            return None
+        if isinstance(value, int) and value > 0:
+            return value
+        if isinstance(value, str) and value.isdigit():
+            parsed_value = int(value)
+            if parsed_value > 0:
+                return parsed_value
+        return None
+
     def _context_length_from_scan_payload(self, model_payload: dict[str, typing.Any] | None) -> int | None:
         if not model_payload:
             return None
 
         for field_name in ('context_length', 'context_window', 'max_context_length'):
-            value = model_payload.get(field_name)
+            context_length = self._positive_int(model_payload.get(field_name))
-            if isinstance(value, bool):
+            if context_length is not None:
-                continue
+                return context_length
-            if isinstance(value, int) and value > 0:
-                return value
-            if isinstance(value, str) and value.isdigit():
-                parsed_value = int(value)
-                if parsed_value > 0:
-                    return parsed_value
         return None
 
+    def _context_length_from_litellm_model_info(self, model_info: typing.Any) -> int | None:
+        if isinstance(model_info, dict):
+            return self._positive_int(model_info.get('max_input_tokens'))
+        return self._positive_int(getattr(model_info, 'max_input_tokens', None))
+
     def _metadata_provider_candidates(self, model_name: str) -> list[str]:
         normalized_model_name = (model_name or '').lower()
         candidates = []
@@ -126,7 +137,7 @@ class LiteLLMRequester(requester.ProviderAPIRequester):
         return None
 
     def _safe_context_length(self, model_name: str) -> int | None:
-        helper = getattr(litellm, 'get_max_tokens', None)
+        helper = getattr(litellm, 'get_model_info', None)
         if not callable(helper):
             return self._known_context_length_fallback(model_name)
 
@@ -143,11 +154,12 @@ class LiteLLMRequester(requester.ProviderAPIRequester):
                 continue
             tried_candidates.append(candidate)
             try:
-                max_tokens = helper(candidate)
+                model_info = helper(candidate)
             except Exception:
                 continue
-            if isinstance(max_tokens, int) and max_tokens > 0:
+            context_length = self._context_length_from_litellm_model_info(model_info)
-                return max_tokens
+            if context_length is not None:
+                return context_length
         return self._known_context_length_fallback(model_name)
 
     def _supports_function_calling(self, model_name: str) -> bool:

src/langbot/pkg/telemetry/heartbeat.py

@@ -109,6 +109,7 @@ async def build_heartbeat_payload(ap: core_app.Application) -> dict:
         'query_id': '',
         'version': constants.semantic_version,
         'instance_id': constants.instance_id,
+        'instance_create_ts': constants.instance_create_ts,
         'edition': constants.edition,
         'features': features,
         'timestamp': datetime.now(timezone.utc).isoformat(),

src/langbot/pkg/utils/constants.py

@@ -16,3 +16,11 @@ debug_mode = False
 edition = 'community'
 
 instance_id = ''
+
+instance_create_ts = 0
+"""Unix timestamp (seconds) of when this instance was first created.
+
+Sourced from ``data/labels/instance_id.json``. Backfilled to the current
+time for instances created before this field existed, so it is always a
+positive value once load_config has run.
+"""

tests/unit_tests/provider/test_litellmchat.py

@@ -1034,11 +1034,28 @@ class TestScanModels:
             },
         )
 
-        with patch.object(litellmchat.litellm, 'get_max_tokens') as mock_get_max_tokens:
+        with patch.object(litellmchat.litellm, 'get_model_info') as mock_get_model_info:
-            mock_get_max_tokens.side_effect = lambda model: 131072 if model == 'moonshot/moonshot-v1-128k' else None
+            mock_get_model_info.side_effect = (
+                lambda model: {'max_input_tokens': 131072}
+                if model == 'moonshot/moonshot-v1-128k'
+                else {}
+            )
 
             assert requester._safe_context_length('moonshot-v1-128k') == 131072
 
+    def test_safe_context_length_uses_litellm_max_input_tokens(self):
+        """LiteLLM max_output_tokens must not be treated as the context window."""
+        requester = litellmchat.LiteLLMRequester(ap=Mock(), config={})
+
+        with patch.object(litellmchat.litellm, 'get_model_info') as mock_get_model_info:
+            mock_get_model_info.return_value = {
+                'max_input_tokens': 128000,
+                'max_output_tokens': 16384,
+                'max_tokens': 16384,
+            }
+
+            assert requester._safe_context_length('gpt-4o') == 128000
+
     def test_litellm_bool_helper_tries_moonshot_metadata_alias(self):
         """OpenAI-compatible Moonshot endpoints still use Moonshot metadata for abilities."""
         requester = litellmchat.LiteLLMRequester(
@@ -1102,7 +1119,7 @@ class TestScanModels:
             },
         )
 
-        with patch.object(litellmchat.litellm, 'get_max_tokens', side_effect=Exception('not mapped')):
+        with patch.object(litellmchat.litellm, 'get_model_info', side_effect=Exception('not mapped')):
             assert requester._safe_context_length('deepseek-v4-pro') == 1_000_000
             assert requester._safe_context_length('deepseek-v4-flash') == 1_000_000
 

tests/unit_tests/telemetry/test_heartbeat.py

@@ -62,6 +62,7 @@ class TestBuildHeartbeatPayload:
 
         assert payload['event_type'] == 'instance_heartbeat'
         assert payload['query_id'] == ''
+        assert 'instance_create_ts' in payload
         assert 'timestamp' in payload
         f = payload['features']
         assert f['database'] == 'postgresql'

uv.lock

@@ -1967,7 +1967,7 @@ wheels = [
 
 [[package]]
 name = "langbot"
-version = "4.10.1"
+version = "4.10.2"
 source = { editable = "." }
 dependencies = [
     { name = "aiocqhttp" },