Ln 514 Test Log Analyzer

Points your LLM at application logs and gets back a classification of what actually matters. Splits errors into four buckets (real bugs, test artifacts, expected behavior, operational warnings) so you know what to fix versus what to ignore. Also audits log quality itself: checks if you're logging at the wrong levels, missing trace IDs, or spamming DEBUG lines in production. The two-layer design is smart: a Node script does the collection and counting, Claude does the judgment calls. Works with Docker Compose logs, flat files, or Loki. Useful after test runs when you need to separate signal from noise, or when your logs have grown into a mess and you want an assessment of what needs cleanup.

Install to Claude Code

npx -y skills add levnikolaevich/claude-code-skills --skill ln-514-test-log-analyzer --agent claude-code