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.
mdr add dayfinggg/openai-codex-agent-skills/reflect@git:20260905.b1dcd2amdr add dayfinggg/openai-codex-agent-skills/reflect@sha256:ec10754217ba2ce9[](https://markdownregistry.com/a/art_5vauhvdkntxxdbvt)
0 badge views in 30 days
| version | committed | commit | size | audit | |
|---|---|---|---|---|---|
| git:20260905.b1dcd2a latest | 2026-09-05 | b1dcd2a | 2,065 B | A | view · diff |
| git:20260830.cc6c9ed | 2026-08-30 | cc6c9ed | 1,583 B | A | view |
dayfinggg/openai-codex-agent-skills · 4 stars · license MIT · pushed 2026-09-05 · branch main
GET https://markdownregistry.com/api/v1/artifacts/art_5vauhvdkntxxdbvt GET https://markdownregistry.com/api/v1/resolve?ref=dayfinggg/openai-codex-agent-skills/reflect GET https://markdownregistry.com/api/v1/blob/ec10754217ba2ce940a736eb3e6142dfbbca9cbe3c3eae3bda8b97babb0f2ce1