fix: provider/model filters were silently ignored + startup blocked HTTP for the full reindex

stackunderflow/routes/cost.py

@@ -16,9 +16,9 @@
 from __future__ import annotations
 
 from pathlib import Path
-from typing import Any
+from typing import Annotated, Any
 
-from fastapi import APIRouter, HTTPException
+from fastapi import APIRouter, HTTPException, Query
 
 import stackunderflow.deps as deps
 from stackunderflow.infra.currency import active_currency_payload
@@ -153,7 +153,7 @@ async def get_cost_data(log_path: str | None = None, timezone_offset: int = 0):
 @router.get("/api/cost-data/by-provider")
 async def get_cost_by_provider(
     period: str = "month",
-    provider: list[str] | None = None,
+    provider: Annotated[list[str] | None, Query()] = None,
 ):
     """Return total cost / message count / session count grouped by provider.
 

stackunderflow/routes/data.py

@@ -5,8 +5,9 @@
 import threading
 import time
 from pathlib import Path
+from typing import Annotated
 
-from fastapi import APIRouter, HTTPException
+from fastapi import APIRouter, HTTPException, Query
 from fastapi.responses import JSONResponse
 
 import stackunderflow.deps as deps
@@ -113,8 +114,8 @@ async def get_stats(timezone_offset: int = 0):
 @router.get("/api/dashboard-data")
 async def get_dashboard_data(
     timezone_offset: int = 0,
-    provider: list[str] | None = None,
-    model: list[str] | None = None,
+    provider: Annotated[list[str] | None, Query()] = None,
+    model: Annotated[list[str] | None, Query()] = None,
 ):
     """Get optimized data for initial dashboard load.
 
@@ -278,8 +279,8 @@ def _apply_currency_to_stats(stats: dict) -> dict:
 async def get_messages(
     limit: int | None = None,
     timezone_offset: int = 0,
-    provider: list[str] | None = None,
-    model: list[str] | None = None,
+    provider: Annotated[list[str] | None, Query()] = None,
+    model: Annotated[list[str] | None, Query()] = None,
 ):
     """Get messages for the current project.
 

stackunderflow/routes/optimize.py

@@ -14,7 +14,9 @@
 
 from __future__ import annotations
 
-from fastapi import APIRouter, HTTPException
+from typing import Annotated
+
+from fastapi import APIRouter, HTTPException, Query
 
 import stackunderflow.deps as deps
 from stackunderflow.reports.optimize import find_patterns, find_waste
@@ -30,8 +32,8 @@
 @router.get("/api/optimize")
 async def get_optimize_report(
     period: str = "30days",
-    project: list[str] | None = None,
-    exclude: list[str] | None = None,
+    project: Annotated[list[str] | None, Query()] = None,
+    exclude: Annotated[list[str] | None, Query()] = None,
 ):
     """Run waste + structural-pattern detection over *period*.
 

stackunderflow/routes/projects.py

@@ -3,8 +3,9 @@
 import os
 from collections import defaultdict
 from pathlib import Path
+from typing import Annotated
 
-from fastapi import APIRouter, HTTPException
+from fastapi import APIRouter, HTTPException, Query
 from fastapi.responses import JSONResponse
 
 import stackunderflow.deps as deps
@@ -155,7 +156,7 @@ async def get_projects(
     sort_by: str = "last_modified",
     limit: int | None = None,
     offset: int = 0,
-    provider: list[str] | None = None,
+    provider: Annotated[list[str] | None, Query()] = None,
 ):
     """
     Get all available Claude projects with metadata.

stackunderflow/routes/sessions.py

@@ -3,8 +3,9 @@
 import json
 from datetime import datetime
 from pathlib import Path
+from typing import Annotated
 
-from fastapi import APIRouter, HTTPException
+from fastapi import APIRouter, HTTPException, Query
 from fastapi.responses import JSONResponse
 
 import stackunderflow.deps as deps
@@ -38,7 +39,7 @@ def _duration_minutes(first: str | None, last: str | None) -> float | None:
 @router.get("/api/jsonl-files")
 async def get_jsonl_files(
     project: str | None = None,
-    provider: list[str] | None = None,
+    provider: Annotated[list[str] | None, Query()] = None,
 ):
     """Get list of JSONL files for a project with metadata.
 

stackunderflow/server.py

@@ -89,20 +89,40 @@ async def _lifespan(_app: FastAPI):
     if failed:
         logger.warning(f"Failed services: {', '.join(failed)}")
 
-    # Initialise the session store and run one ingest pass.
+    # Initialise the session store schema synchronously (cheap), then
+    # run the ingest in a background thread so HTTP starts serving
+    # immediately. Without this, the lifespan blocks the bind for the
+    # full duration of the reindex (~90s on 7 small projects, 30+min
+    # on a cold 188-project store) — and the "live at..." line that
+    # already printed from the CLI wrapper is misleading because the
+    # HTTP server hasn't actually started yet.
+    import threading
     from stackunderflow.adapters import registered
     from stackunderflow.ingest import run_ingest
     from stackunderflow.store import db, schema
 
     try:
-        store_conn = db.connect(deps.store_path)
-        schema.apply(store_conn)
-        counts = run_ingest(store_conn, registered())
-        logger.info("Ingest complete: %s", counts)
-        store_conn.close()
-        _maybe_clean_cold_cache()
+        _schema_conn = db.connect(deps.store_path)
+        schema.apply(_schema_conn)
+        _schema_conn.close()
     except Exception as e:
-        logger.error("Ingest failed at startup: %s", e)
+        logger.error("Schema apply failed at startup: %s", e)
+
+    def _background_ingest() -> None:
+        try:
+            conn = db.connect(deps.store_path)
+            counts = run_ingest(conn, registered())
+            logger.info("Ingest complete: %s", counts)
+            conn.close()
+            _maybe_clean_cold_cache()
+        except Exception as exc:  # noqa: BLE001 — top of background thread
+            logger.error("Background ingest failed: %s", exc)
+
+    threading.Thread(
+        target=_background_ingest,
+        name="stackunderflow-ingest",
+        daemon=True,
+    ).start()
 
     yield