Reflect

Extract reusable, evidence-supported improvements after a completed engineering session, repeated failure, or explicit retrospective request. Use when the goal is to improve future agent behavior or an existing workflow. Do not activate automatically after every task or create new rules from a single preference.

dayfinggg d59d54f 2 files · 2.3 KB Updated

File contents

dayfinggg/openai-codex-agent-skills/tree/main/codex/skills/reflect commit d59d54f680

Frequently asked questions

npx skillmds@latest add dayfinggg/reflect