OpenSkillIndex← back to search
claude-code

Awesome-omni-skill observability-instrumentation

Use when adding endpoints/background jobs/integrations where telemetry matters. Do NOT refactor unrelated code. Prefer OpenTelemetry-friendly patterns.

install
source · Clone the upstream repo
git clone https://github.com/diegosouzapw/awesome-omni-skill
Claude Code · Install into ~/.claude/skills/
T=$(mktemp -d) && git clone --depth=1 https://github.com/diegosouzapw/awesome-omni-skill "$T" && mkdir -p ~/.claude/skills && cp -r "$T/skills/development/observability-instrumentation" ~/.claude/skills/diegosouzapw-awesome-omni-skill-observability-instrumentation && rm -rf "$T"
manifest: skills/development/observability-instrumentation/SKILL.md
source content

Principles:

  • Low-cardinality metrics labels.
  • Structured logs with correlation IDs; never log secrets.
  • Trace the "edges": inbound request, outbound calls, key domain operations.

Checklist:

  • Tracing
    • Create spans around external calls (HTTP, DB, message bus, 3rd party SDK)
    • Add useful span attributes (route, status code, dependency name)
    • Propagate context across async boundaries and between services
  • Metrics
    • Add counters for throughput and errors
    • Add histograms for latency where needed
    • Avoid user IDs, request IDs, or other high-cardinality labels
  • Logging
    • Structured logging (message templates)
    • Correlation/trace IDs included (or derived from OTel)
    • Redaction of PII/secrets
  • Health & readiness
    • Ensure health endpoints or probes reflect real dependencies (as per repo convention)
  • Verification
    • Verify local export path (console/OTLP) and sample traces appear
    • Add minimal docs: how to run locally + where to look in dashboards

Finish with:

  • What telemetry was added (spans/metrics/logs)
  • How to validate locally
  • Risks (cardinality, PII, perf)