Skip to content
Get packs

Refit: Environment Retrospective

Turns recorded agent friction into approved fixes to checks, docs, tooling, or access.

Source Download skill
Repository
Yeachan-Heo/oh-my-claudecode
Stars
40,512
Licence
MIT
Files
1 file

What it does

Refit looks back at the working environment an AI coding agent used, not at the code it wrote. It reads OMC's own logs of where sessions stalled, repeated steps, or hit errors, then sorts each problem into the place that should fix it: an automated check, a steering doc, a tool or config, or access to missing information. Nothing is changed until you approve each proposed item, and it can state plainly when no real problems were found.

Great for

  • Developers who use OMC on a repo and want recurring agent friction reduced over time
  • Teams that want their lint, CI, or hook setup audited against real session problems
  • Maintainers who run OMC sessions often and keep plan notepads of accumulated issues

Skip it if

  • Reviewing the correctness of code the agent produced; the skill says review and verify handle that
  • Repos with no OMC session history, since there are no instruments to read and the result will be empty
  • Unattended use without reading the output: headless mode writes proposal files and can be scheduled by cron or CI, so proposals should be checked before any approval

Try asking

  • “Run a refit to see what friction the last few autopilot sessions recorded and where each fix should go”
  • “Survey my ralph loop logs and tell me which checks keep failing and whether they should be a lint rule or CI job”
  • “Do a refit on this repo and show me the proposals before changing anything”
  • “Check the pending refit proposals from last night's headless run and approve the ones that make sense”
Step-by-step guide: how to use Refit: Environment RetrospectiveInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 1 original files, the MIT licence and an attribution note.

    refit.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip refit.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip refit.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Read the evidence line on each proposal before approving; it shows which instrument produced the finding
  • Approve or veto proposals individually rather than in bulk
  • Expect a new deterministic check to be proven by a deliberate failing run before it counts as landed

What's inside

  • SKILL.md

Redistributed unmodified from Yeachan-Heo/oh-my-claudecode under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.