Stats
Actions
Tags
From rune
Post-deploy monitoring agent for HTTP health checks, response time measurement (<500ms FAST, 500-2000ms ACCEPTABLE, >2000ms SLOW), and error detection (4xx/5xx, timeouts). Verifies live sites invoked by deploy/launch. Precision over recall.
How this agent operates — its isolation, permissions, and tool access model
Agent reference
rune:agents/watchdoghaikuThe summary Claude sees when deciding whether to delegate to this agent
You are the **watchdog** skill — Rune's deployment health monitor. **Workflow:** 1. Accept base_url + endpoints (default: `/`, `/health`, `/ready`) 2. HTTP status checks: classify 2xx/3xx/4xx/5xx 3. Response time: <500ms=FAST, 500-2000ms=ACCEPTABLE, >2000ms=SLOW 4. Detect errors: 4xx/5xx, timeouts, empty responses 5. Performance patterns: consistent slowness, cluster degradation, spikes 6. Repo...
You are the watchdog skill — Rune's deployment health monitor.
Workflow:
/, /health, /ready)Critical Rules:
Read skills/watchdog/SKILL.md for the full specification.
npx claudepluginhub rune-kit/rune --plugin @rune/analyticsRuns post-deploy canary monitoring on production URLs: health checks, regression detection for response time/status changes/error patterns/content, optional browser verification. Assesses health and recommends actions like rollback.
Lightweight early warning system for spot-checking critical services (dashboard, wordblocks etc.), infrastructure (Caddy, disk, memory), DBs, and recent changes. Generates quick Markdown status reports in <60s.
Delegate to this agent for performance optimization, load testing (k6/Locust/JMeter), profiling bottlenecks, observability (Prometheus/Grafana/OpenTelemetry), monitoring setup, database tuning, caching (Redis), and capacity planning.