feat(lens): always read Lens traces from Postgres spend logs

Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>

litellm/proxy/proxy_server.py

@@ -1961,12 +1961,11 @@ async def _proxy_lifespan(app: FastAPI, resources: AsyncExitStack) -> AsyncGener
         TypeAdapter(dict[str, object] | None).validate_python(general_settings.get("tracing")),
     )
     tracing_enabled: Final = is_lens_tracing_enabled(tracing_settings)
-    async with manage_tracing(
-        enabled=tracing_enabled,
-        settings=tracing_settings,
-        fallback_factory=lambda: _spend_log_trace_receiver(database),
-    ) as receiver:
-        state: Final[ProxyLifespanState] = {"tracing_receiver": receiver, "database": database}
+    async with manage_tracing(enabled=tracing_enabled, settings=tracing_settings):
+        state: Final[ProxyLifespanState] = {
+            "tracing_receiver": _spend_log_trace_receiver(database),
+            "database": database,
+        }
         from litellm.proxy.admin_mcp import admin_mcp_lifespan
 
         telemetry_settings: Final = load_telemetry_settings()

litellm/proxy/tracing_runtime.py

@@ -40,10 +40,9 @@ async def manage_tracing(
     receiver_factory: Callable[[], TraceReceiver] | None = None,
     settings: Mapping[str, object] | None = None,
     client_factory: Callable[[LensConnection], httpx.AsyncClient] = LensConnection.lifespan_client,
-    fallback_factory: Callable[[], TraceReceiver | None] = lambda: None,
 ) -> AsyncGenerator[TraceReceiver | None, None]:
     if not enabled:
-        yield fallback_factory()
+        yield None
         return
     try:
         connection: Final = LensConnection.from_env()

litellm/tracing/AGENTS.md

@@ -1,7 +1,7 @@
 # Lens gateway boundary
 
 - Lens owns trace normalization, schemas, storage, graph assembly and querying in `BerriAI/lens`
-- This package owns the HTTP client, gateway response validation and asynchronous gateway spend export. Keep the trace reader remote-only
+- This package owns the HTTP client, gateway response validation, asynchronous gateway spend export and the trace reader. Trace reads always come from Postgres spend logs in `spend_logs.py`, even when a Lens service is configured, so traces have one source of truth
 - Gateway endpoints preserve authenticated user/team scope, trace references, pagination and public error categories
 - `generated/` comes from `scripts/generate_trace_types.py` using the Lens schemas pinned in `scripts/lens_assets/source.json`. Update the canonical Lens Rust contracts and import their schema outputs before regenerating. Never edit generated Python
 - Keep the pinned schema and fixture checksums current. CI verifies assets and regenerated Python without a Rust trace implementation in this repository