What is considered a P0 priority finding?

P0 findings are critical production blind spots, such as having no error tracking installed or missing health endpoints, which prevent basic uptime monitoring.

Does this skill automatically fix the issues it finds?

No, this is a diagnostic primitive. It identifies and reports issues. You should use /fix-observability or /log-observability-issues to act on the findings.

What core areas does the observability audit cover?

The skill audits four primary areas: Error Tracking (Sentry), Logging (Structured logging vs console), Health Endpoints (API status and depth), and Analytics (PostHog configuration).

Why is PostHog specifically recommended over other analytics tools?

The skill prioritizes PostHog because it offers API, CLI, and MCP access, allowing Claude and other AI tools to interact with the data programmatically, unlike Vercel Analytics.

Check Observability

Name: Check Observability
Author: phrazzld

byphrazzld

•

Analytics & Monitoring

Audits production observability infrastructure including error tracking, structured logging, health checks, and product analytics.

Check Observability is a diagnostic primitive designed to identify critical blind spots in your application's monitoring stack. It systematically inspects your codebase for Sentry configurations, validates the depth of health check endpoints, checks for structured logging libraries like Pino or Winston, and ensures PostHog is correctly implemented for analytics. By generating a prioritized report of findings ranging from P0 (critical gaps) to P3 (minor improvements), it provides a clear roadmap for improving production visibility and ensuring that your application is ready for reliable operation.

Key Features

015 GitHub stars

02Detection of structured logging vs. excessive console statement usage

03Deep health check analysis for database and service dependencies

04PostHog integration verification for AI-compatible analytics workflows

05Silent failure detection for empty catch blocks and swallowed errors

06Automated Sentry configuration and DSN validation

Use Cases

01Identifying why errors are going uncaptured in production environments

02Conducting pre-launch production readiness audits

03Standardizing observability patterns across a team's codebase

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add phrazzld/claude-config check-observability

For use in Claude.ai and ChatGPT

Download Skill