ln-627-observability-auditor_skill

Overview

This skill audits observability for a codebase, covering structured logging, health check endpoints, metrics collection, request tracing, and log levels. It produces a findings report with severity, location, estimated effort, and concrete recommendations, plus a numeric compliance score. The worker is framework-aware and returns a concise summary for coordinator pipelines.

How this skill works

The auditor parses the provided context (tech stack, framework, code root, and output directory) and scans source files for observability patterns. It applies rule sets to detect unstructured logging, missing health endpoints, absent metrics, lack of request tracing, and improper log level usage. Findings are collected, scored per the shared scoring formula, written as a single report file to the output directory, and a minimal summary is returned to the coordinator.

When to use it

• As part of a delivery pipeline to validate runtime visibility before release
• During architecture or QA gates to identify observability blind spots
• When onboarding a new service or refactoring to ensure instrumentation standards
• Before deploying to production or scaling services in distributed environments
• As a recurring audit to track observability drift over time

Best practices

• Run the audit with accurate context (framework and runtime) to enable framework-aware detections
• Exclude test and dev-only scripts from console.log checks to avoid false positives
• Prioritize HIGH findings (e.g., missing health endpoints) for immediate remediation
• Adopt structured loggers and consistent log levels across services
• Instrument minimal Prometheus metrics and request ID propagation early to enable distributed debugging

Example use cases

• CI pipeline audit step that fails the release on LOW compliance score
• Pre-release checklist item to ensure /health and readiness endpoints exist
• Refactor validation: confirm console.log calls were replaced with structured logger usage
• Observability maturity check during architecture reviews to estimate remediation effort
• Post-incident audit to identify missing tracing or metrics that hindered root cause analysis