[BUG] scarf.finding() + posthog.finding() block asyncio event loop up to 20s per finding

[chirag127]

Problem

create_vulnerability_report (async function_tool) reaches posthog.finding() and scarf.finding() on the event loop; both use blocking urllib.request.urlopen(..., timeout=10). When either endpoint is slow/unreachable, each finding stalls the loop up to ~20s combined, freezing every other agent tool call in the scan.

Call path

• strix/tools/reporting/tool.py:261 — report_state.add_vulnerability_report(...) called synchronously from async def _do_create.
• strix/report/state.py:269-270 — calls posthog.finding(severity) then scarf.finding(severity).
• strix/telemetry/posthog.py:45-46 — urllib.request.urlopen(req, timeout=10).
• strix/telemetry/scarf.py:46-47 — same pattern, timeout=10.

Expected

Telemetry sends off the event loop (e.g. asyncio.to_thread, background queue, or httpx.AsyncClient).

Actual

Each finding blocks the event loop up to ~20s cumulative when telemetry endpoints are slow.

Environment

strix-agent 1.0.4, Python >=3.12 (per pyproject.toml).

Thanks for maintaining usestrix/strix!

[pvratnam203-dev]

import asyncio
import urllib.request

... existing code ...

def _send_request(req):
"""Execute the blocking urlopen request."""
try:
with urllib.request.urlopen(req, timeout=10) as response:
return response.status
except Exception:
pass

async def finding(severity: str):
"""Send telemetry event off the main asyncio event loop."""
# ... construct your request (req) here ...

# Run the blocking urlopen in a separate thread
await asyncio.to_thread(_send_request, req)