8b1ce13da7
argus: add Phase 1 active monitoring system
...
- What: alerting.py (7 health checks), alerting_routes.py (3 endpoints),
PATCH_INSTRUCTIONS.md (app.py integration guide for Rhea)
- Why: engineering acceleration initiative — move from passive dashboard
to active monitoring with agent health, quality regression, throughput
anomaly, stuck loop, cost spike, and domain rejection pattern detection
- Endpoints: GET /check, GET /api/alerts, GET /api/failure-report/{agent}
- Deploy: Rhea applies PATCH_INSTRUCTIONS to live app.py, restarts service,
adds 5-min systemd timer for /check
Pentagon-Agent: Argus <9aa57086-bee9-461b-ae26-dfe5809820a8>
2026-04-14 18:14:07 +00:00
9fea4fc7df
Delete stale diagnostics/ report files — data lives in pipeline.db
...
Removed pr-log.md (1,211 entries from Mar 25), evolution.md, weekly report,
and PATCH_INSTRUCTIONS.md (superseded by ops/AGENT-SOP.md). All regenerable
from pipeline.db on demand.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-14 16:19:24 +01:00
0d718f0786
Session capture: 20260414-114853
2026-04-14 12:48:53 +01:00
4e20986c25
ship: add agent SOP, auto-deploy infrastructure, cleanup stale files
...
- AGENT-SOP.md: enforceable checklist for commit/review/deploy cycle
- auto-deploy.sh + systemd units: 2-min timer pulls from Forgejo, syncs
to working dirs, restarts services only when Python changes, smoke tests
- prune-branches.sh: dry-run-by-default branch cleanup tool
- Delete root diagnostics/ (stale artifacts, all code moved to ops/)
- Delete 7 orphaned HTML prototypes (untracked, local-only)
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-14 12:46:25 +01:00
33e670b436
argus: add active alerting system (Phase 1)
...
Three new files for the engineering acceleration initiative:
- alerting.py: 7 health check functions (dormant agents, quality regression,
throughput anomaly, rejection spikes, stuck loops, cost spikes, domain
rejection patterns) + failure report generator
- alerting_routes.py: /check, /api/alerts, /api/failure-report/{agent} endpoints
- PATCH_INSTRUCTIONS.md: integration guide for app.py (imports, route
registration, auth middleware bypass, DB connection)
Observe and alert only — no pipeline modification. Independence constraint
is load-bearing for measurement trustworthiness.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-03-28 22:45:07 +00:00
