Observability
AI-agent skills for software engineering — works with Claude Code, Codex, Cursor
npx -y skills add vaibhavsaxena022/skills --skill observabilityAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
2 things to look at
- 23 days oldThe repository was created 23 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
- 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
Add production observability to a service. Use when a service is hard to debug in prod or lacks logging/metrics/tracing. Adds structured logs, key metrics, and trace context — without noise.
SKILL.md
1.1 KB, as published. Nobody here has run it
Observability
Make the service debuggable in production.
Add
- Structured logs — key events at the right level, each carrying a correlation/trace id. No secrets/PII. No logging in hot loops.
- Metrics — RED for endpoints (Rate, Errors, Duration) plus key business counters; histograms for latency.
- Tracing — propagate a trace/correlation id across the service and its external calls.
- Health — liveness and readiness endpoints.
Rules
- Log intent and outcomes, not noise — every line should be actionable.
- Prefer metrics over log-scraping for rates and latencies.
- Thread one correlation id through a request end-to-end.
- Never log request bodies containing sensitive data.