Agent session replay and traces. Free $0. Pro is $40/month. Enterprise is custom. GitHub is about 4.4k stars, not 5k+ unspecified.

Superpowers' debugging skill forces coding agents through four phases: root cause, pattern analysis, one hypothesis at a time, then a tested fix.

AWS's AI-first observability experience: one workspace to trace, evaluate, and troubleshoot applications and agents across accounts, regions, and clouds.

Andon is a MIT Lean quality kit for Claude Code that turns each agent defect into a permanent countermeasure.

Reference materials for Makepad development including troubleshooting common errors, code quality guidelines, and responsive layout patterns for cross-platform apps.