fix(cli): bump --remote query timeout to 300s, add AGNES_QUERY_TIMEOUT

The httpx client behind 'agnes query --remote' used the default 30s
timeout, killing every BigQuery SELECT that took longer than half a
minute — i.e. most non-trivial remote queries.

cli/client.py now exposes QUERY_TIMEOUT_S (default 300s, override via
AGNES_QUERY_TIMEOUT) and propagates a kw-only 'timeout' through
api_get/post/delete/patch. _query_remote passes QUERY_TIMEOUT_S so only
the long-running /api/query path gets the bump; every other CLI call
keeps the 30s default.

Server-side has no read deadline on /api/query, so the client cap was
the sole bottleneck.

CHANGELOG.md

@@ -10,6 +10,10 @@ CalVer image tags (`stable-YYYY.MM.N`, `dev-YYYY.MM.N`) are produced for every C
 
 ## [Unreleased]
 
+### Fixed
+
+- `agnes query --remote` no longer dies after 30s on long-running BigQuery SELECTs. The CLI HTTP client now defaults to a 300s timeout for `/api/query` and exposes `AGNES_QUERY_TIMEOUT` (seconds, float) for operators who need to extend it further. Other CLI calls keep the 30s default. (`cli/client.py`, `cli/commands/query.py`)
+
 ## [0.35.0] — 2026-05-05
 
 Five-defect fix for the silently-broken session pipeline on default Compose deploys (#176). Sessions uploaded by `agnes push` landed on `/data/user_sessions/<user>/*.jsonl`, but on a stock `docker compose up` deploy nothing ever processed them — `/corporate-memory` stayed empty even when sessions and `CLAUDE.local.md` were uploaded. The root cause was a stack of compounding defects: LLM SDKs were dev-only deps so the scheduler container boot-looped on `ModuleNotFoundError`, the side-car services were profile-gated and ran as tight `restart: unless-stopped` boot loops anyway, the `verification_detector` had no scheduler entry at all, the first-time setup never seeded an `ai:` block, and the `/corporate-memory` page silently filtered out the pending review queue. This release wires the LLM pipeline into the existing scheduler-v2 model (one HTTP-driven cron tick per service) and adds a health-check that warns when uploaded jsonls aren't being processed.

cli/client.py

@@ -15,6 +15,11 @@ from cli.config import get_server_url, get_token
 _RETRY_ATTEMPTS = int(os.environ.get("AGNES_STREAM_RETRIES", "3"))
 _RETRY_BACKOFFS_S = (0.3, 1.0, 3.0)  # seconds before attempt 2, 3, 4
 
+# Long-running query timeout. /api/query forwards to BigQuery for remote
+# tables, where SELECTs routinely run for minutes. The default 30s HTTP
+# timeout dies long before BQ finishes. Operators tune via AGNES_QUERY_TIMEOUT.
+QUERY_TIMEOUT_S = float(os.environ.get("AGNES_QUERY_TIMEOUT", "300"))
+
 
 def get_client(timeout: float = 30.0) -> httpx.Client:
     """Get an authenticated httpx client."""
@@ -29,23 +34,23 @@ def get_client(timeout: float = 30.0) -> httpx.Client:
     )
 
 
-def api_get(path: str, **kwargs) -> httpx.Response:
-    with get_client() as client:
+def api_get(path: str, *, timeout: float = 30.0, **kwargs) -> httpx.Response:
+    with get_client(timeout=timeout) as client:
         return client.get(path, **kwargs)
 
 
-def api_post(path: str, **kwargs) -> httpx.Response:
-    with get_client() as client:
+def api_post(path: str, *, timeout: float = 30.0, **kwargs) -> httpx.Response:
+    with get_client(timeout=timeout) as client:
         return client.post(path, **kwargs)
 
 
-def api_delete(path: str, **kwargs) -> httpx.Response:
-    with get_client() as client:
+def api_delete(path: str, *, timeout: float = 30.0, **kwargs) -> httpx.Response:
+    with get_client(timeout=timeout) as client:
         return client.delete(path, **kwargs)
 
 
-def api_patch(path: str, **kwargs) -> httpx.Response:
-    with get_client() as client:
+def api_patch(path: str, *, timeout: float = 30.0, **kwargs) -> httpx.Response:
+    with get_client(timeout=timeout) as client:
         return client.patch(path, **kwargs)
 
 

cli/commands/query.py

@@ -85,10 +85,14 @@ def _query_local(sql: str, fmt: str, limit: int):
 
 def _query_remote(sql: str, fmt: str, limit: int):
     """Run query against server DuckDB via API."""
-    from cli.client import api_post
+    from cli.client import QUERY_TIMEOUT_S, api_post
     from cli.error_render import render_error
 
-    resp = api_post("/api/query", json={"sql": sql, "limit": limit})
+    resp = api_post(
+        "/api/query",
+        json={"sql": sql, "limit": limit},
+        timeout=QUERY_TIMEOUT_S,
+    )
     if resp.status_code != 200:
         # Parse JSON body if possible, fall back to text. The shared
         # renderer pretty-prints typed BQ errors (cross_project_forbidden,

tests/test_cli_query.py

@@ -55,6 +55,23 @@ class TestRemoteQuery:
         assert result.exit_code == 0
         assert "truncated" in result.output
 
+    def test_remote_query_uses_long_timeout(self):
+        """--remote passes the long-running QUERY_TIMEOUT_S to api_post.
+
+        BigQuery SELECTs routinely take minutes; the default 30s httpx
+        timeout dies long before the query finishes. Regression guard for
+        the fix that introduced AGNES_QUERY_TIMEOUT (default 300s).
+        """
+        from cli.client import QUERY_TIMEOUT_S
+
+        payload = {"columns": [], "rows": [], "truncated": False}
+        mock_post = MagicMock(return_value=_resp(200, payload))
+        with patch("cli.client.api_post", mock_post):
+            result = runner.invoke(app, ["query", "SELECT 1", "--remote"])
+        assert result.exit_code == 0
+        assert mock_post.call_args.kwargs["timeout"] == QUERY_TIMEOUT_S
+        assert QUERY_TIMEOUT_S >= 300.0
+
 
 class TestLocalQuery:
     def test_local_query_no_db(self, tmp_config):