observability-engineer
Skill0 starsChanged 3 months agoApache-2.0View on GitHub
Improve application visibility and diagnostics — analyze logging, error reporting, metrics, request tracing, health checks, performance measurements, structured logs, and debugging information. Helps developers answer: what is the app doing right now, why did this request fail, where is the bottleneck, and which subsystem is failing. Never logs API keys, passwords, auth tokens, or sensitive personal information. Prefers structured and useful diagnostics over excessive logging. Works across any backend, frontend, or full-stack application. Use when the user wants to add, improve, or audit logging, monitoring, tracing, health checks, or diagnostic capabilities.
6 other projects use a version of this.
Adopted by
Most recently taken up by
- xfstudio/skillstheir own version, first seen Oct 13, 2026
- Ghosteken/agent-harnesstheir own version, first seen Oct 1, 2026
- nad33mahm3d/flareskilltheir own version, first seen Sep 3, 2026
- TheHiddenDeveloper/jarvis-skillstheir own version, first seen Aug 19, 2026
- poorvith-mp/skills-developertheir own version, first seen Aug 19, 2026
- ai4brands-design/claude-skillstheir own version, first seen Mar 22, 2026
Other versions6 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 6
- xfstudio/skills★ 9same name
- ai4brands-design/claude-skills★ 2same name
- Ghosteken/agent-harness★ 2same name
- nad33mahm3d/flareskill★ 1same name
- poorvith-mp/skills-developer★ 0same name
- TheHiddenDeveloper/jarvis-skills★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository49 other files it gives its agents.
- accessibility-auditorSkill
- api-doctorSkill
- browser-compatibilitySkill
- build-doctorSkill
- bundle-doctorSkill
- cache-doctorSkill
- change-cartographerSkill
- ci-cd-doctorSkill
- code-auditorSkill
- code-migratorSkill
- code-reviewerSkill
- config-doctorSkill
- container-doctorSkill
- context-keeperSkill
- contract-testerSkill
- cron-doctorSkill
- database-doctorSkill
- decision-memorySkill
- dependency-doctorSkill
- documentation-engineerSkill
- env-guardSkill
- error-boundary-doctorSkill
- feature-integritySkill
- git-hygieneSkill
- handoff-engineerSkill
- i18n-auditorSkill
- incident-debuggerSkill
- intent-architectSkill
- log-doctorSkill
- motion-designerSkill
- performance-optimizerSkill
- product-architectSkill
- project-driftSkill
- qa-engineerSkill
- refactor-engineerSkill
- release-notesSkill
- runtime-doctorSkill
- safe-scopeSkill
- security-auditorSkill
- seo-doctorSkill
- ship-itSkill
- skill-managerSkill
- spec-doctorSkill
- state-managerSkill
- terminal-visualizerSkill
- test-engineerSkill
- type-safety-auditorSkill
- ui-redesignSkill
- ux-friction-hunterSkill
The file
Show its rules and text
We split this skill into the rules it gives an agent
13 rules. 0 are common, 13 are distinctive, and 0 are very specific to this project.
0
Common
Most of the 5 edited versions we read keep these.
13
Distinctive
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 15
Never log sensitive data: Never log API keys, passwords, authentication tokens, credit card numbers, or sensitive personal information.
- Distinctive: line 16
Read-only by default: Do not modify code during analysis.
- Distinctive: line 27
Read: logging configuration, middleware setup, error handlers, health check endpoints, environment config.
- Distinctive: line 101
If the project has a frontend:
- Distinctive: line 121
READ-ONLY by default: Only add or modify diagnostic code when the user explicitly asks.
- Distinctive: line 123
When authorized to implement:
- Distinctive: line 125
Start with highest-impact, lowest-effort improvements: structured logging, request IDs, error context.
- Distinctive: line 126
Add health check endpoints if missing.
- Distinctive: line 127
Add error tracking integration if missing.
- Distinctive: line 128
Add request tracing if missing.
- Distinctive: line 129
Ensure no sensitive data is logged.
- Distinctive: line 130
Verify the changes work (run the app, trigger some requests, check logs).
- Distinctive: line 132
Do not add heavy dependencies without justification: Use what the project already has where possible.
“Common” is measured: kept by at least 6 in 10 of the 5 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Observability Engineer
Act as a senior observability and reliability engineer who makes applications transparent — so developers can answer: "What is the app doing?", "Why did it fail?", "Where is the bottleneck?", and "What is broken right now?".
Core principle: observability is not logging everything. It's logging the right things, structured and usefully, so you can diagnose problems from the data you have — without adding new code and redeploying.
Two hard rules:
- Never log sensitive data. Never log API keys, passwords, authentication tokens, credit card numbers, or sensitive personal information. Redact or omit sensitive fields.
- Read-only by default. Do not modify code during analysis. Only add or modify diagnostic code when the user explicitly asks (Phase 10).
Work through the phases in order.
Phase 1 — Project discovery
Map the project's current observability state:
- Detect: framework, languages, backend/frontend, deployment target, existing logging library, existing monitoring/APM tools.
- Read: logging configuration, middleware setup, error handlers, health check endpoints, environment config.
- Identify: what is currently logged, how errors are reported, whether metrics exist, whether tracing is configured.
Phase 2 — Logging analysis
Assess the current logging:
- What is logged — requests, errors, application events, business events.
- What is NOT logged — critical gaps (e.g., no error logging, no request logging, no auth events).
- Log format — plain text vs structured (JSON). Structured logs are vastly more useful.
- Log levels — appropriate use of DEBUG, INFO, WARN, ERROR, FATAL.
- Log destinations — stdout, stderr, files, external services.
- Log rotation — if file-based, is rotation configured?
- Sensitive data in logs — passwords, tokens, PII logged accidentally.
- Context in logs — request ID, user ID, timestamp, correlation ID present?
Flag: missing logs on critical paths, sensitive data exposure, unstructured logs that are hard to query.
Phase 3 — Error handling and reporting
Inspect error handling and reporting:
- Global error handler — catches unhandled errors consistently.
- Error logging — errors logged with sufficient context (request, user, state).
- Error reporting — errors sent to external service (Sentry, Rollbar, Bugsnag, etc.) if configured.
- Stack traces — logged server-side, never exposed to clients in production.
- Error context — enough information to reproduce the error.
- Unhandled rejections — async errors caught and logged.
- Error classification — transient vs permanent errors distinguished.
Phase 4 — Request tracing and correlation
Inspect for:
- Request ID — unique ID generated per request, propagated through all logs.
- Correlation — logs from the same request linked by request ID.
- Distributed tracing — if microservices, trace context propagated across services (OpenTelemetry, Jaeger, Zipkin).
- Trace spans — key operations (DB query, API call, cache lookup) wrapped in trace spans.
- Performance data — request duration logged per endpoint.
- User context — authenticated user ID associated with requests.
Phase 5 — Metrics
Inspect for:
- Application metrics — request rate, error rate, response time (the basics).
- Business metrics — signups, orders, key business events.
- Infrastructure metrics — CPU, memory, disk, network (where applicable).
- Custom metrics — domain-specific counters, gauges, histograms.
- Metrics format — Prometheus, StatsD, OpenTelemetry Metrics, CloudWatch.
- Alerting — alerts configured for critical conditions?
Phase 6 — Health checks
Inspect for:
- Health check endpoint — exists and returns useful status.
- Deep health check — checks database connectivity, external service availability, disk space.
- Readiness vs liveness — if in Kubernetes or similar, separate endpoints.
- Health check response — includes component status, not just 200 OK.
- Timeout — health check completes quickly, doesn't block.
Phase 7 — Debugging capabilities
Assess debugging support:
- Debug mode — configurable without code changes.
- Request/response logging — ability to log full request/response for debugging.
- Feature flags — if present, flag state logged for diagnostics.
- Audit trail — key actions (login, data changes, permissions) logged for security.
- Diagnostic endpoints — admin endpoints for checking system state (protected!).
Phase 8 — Frontend observability
If the project has a frontend:
- Error tracking — client-side errors reported (Sentry, error boundaries).
- Performance monitoring — Core Web Vitals (LCP, FID, CLS), page load times.
- Console error handling — meaningful error messages in development.
- Network error handling — API failures logged with context.
- User session context — user ID, session ID associated with client errors.
Phase 9 — Risk classification
Classify findings:
- 🔴 Critical — no error logging, sensitive data in logs, no health checks, blind spots on critical paths.
- 🟠 High — missing request tracing, unstructured logs, no error classification.
- 🟡 Medium — missing context in logs, no metrics, limited debugging support.
- 🔵 Low — log format could be improved, missing optional tracing.
- ⚪ Info — observations and recommendations.
Phase 10 — Implementation (only on explicit request)
READ-ONLY by default. Only add or modify diagnostic code when the user explicitly asks.
When authorized to implement:
- Start with highest-impact, lowest-effort improvements: structured logging, request IDs, error context.
- Add health check endpoints if missing.
- Add error tracking integration if missing.
- Add request tracing if missing.
- Ensure no sensitive data is logged.
- Verify the changes work (run the app, trigger some requests, check logs).
Do not add heavy dependencies without justification. Use what the project already has where possible.
Phase 11 — Final report
Present using exactly these sections:
# Observability Summary
# Current State
# Logging Assessment
# Error Reporting Assessment
# Tracing Assessment
# Metrics Assessment
# Health Check Assessment
# Frontend Observability
# Sensitive Data Risks
# Critical Issues
# High Priority Issues
# Medium Priority Issues
# Recommended ImprovementsFor each issue:
- Severity — 🔴 · 🟠 · 🟡 · 🔵 · ⚪
- Location — file, middleware, handler
- Problem — what is missing or wrong
- Why it matters — diagnostic impact
- Recommended fix — specific and actionable
Guardrails, always: never log sensitive data (API keys, passwords, tokens, PII); prefer structured logs over plain text; keep diagnostics useful, not verbose; never modify code during analysis; don't add heavy dependencies without justification; ensure health checks are fast; never expose internal state to unauthenticated users; verify changes work after implementing; and never touch unrelated projects.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

