feat: auto-escalation system + agentic loop fixes (#149) (#149)

Wire up automatic error-to-task escalation and fix the agentic loop
stopping after the first tool call.

Auto-escalation:
- Add swarm.task_queue.models with create_task() bridge to existing
  task queue SQLite DB
- Add swarm.event_log with EventType enum, log_event(), and SQLite
  persistence + WebSocket broadcast
- Wire capture_error() into request logging middleware so unhandled
  HTTP exceptions auto-create [BUG] tasks with stack traces, git
  context, and push notifications (5-min dedup window)

Agentic loop (Round 11 Bug #1):
- Wrap agent_chat() in asyncio.to_thread() to stop blocking the
  event loop (fixes Discord heartbeat warnings)
- Enable Agno's native multi-turn tool chaining via show_tool_calls
  and tool_call_limit on the Agent config
- Strengthen multi-step continuation prompts with explicit examples

Co-authored-by: Trip T <trip@local>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
Alexander Whitestone
2026-03-08 03:11:14 -04:00
committed by GitHub
parent 7792ae745f
commit fd0ede0d51
9 changed files with 323 additions and 12 deletions

View File

@@ -100,13 +100,30 @@ class RequestLoggingMiddleware(BaseHTTPMiddleware):
except Exception as exc:
# Calculate duration even for failed requests
duration_ms = (time.time() - start_time) * 1000
# Log the error
logger.error(
f"[{correlation_id}] {request.method} {request.url.path} "
f"- ERROR - {duration_ms:.2f}ms - {client_ip} - {str(exc)}"
)
# Auto-escalate: create bug report task from unhandled exception
try:
from infrastructure.error_capture import capture_error
capture_error(
exc,
source="http",
context={
"method": request.method,
"path": request.url.path,
"correlation_id": correlation_id,
"client_ip": client_ip,
"duration_ms": f"{duration_ms:.0f}",
},
)
except Exception:
pass # never let escalation break the request
# Re-raise the exception
raise