git-repo-prep
CLAUDE.md2 starsChanged 48 days agoMITView on GitHub
Plugin with two skills for open-source readiness: prepare (full lifecycle) and review (standalone audit). Router in skills/git-repo-prep/SKILL.md dispatches based on user intent. Routes to specific sub-skills based on context: - Preparing for release → git-repo-prep:prepare - Reviewing/auditing → git-repo-prep:review - Ambiguous → AskUserQuestion to clarify When to invoke: - Making a private repo public - Preparing a project for open-source release - Auditing a repo for secrets, missing docs, or openness gaps - Re-checking after making changes (review mode) Core…
In this repository3 other files it gives its agents.
- git-repo-prepSkill
- prepareSkill
- reviewSkill
The file
We split this CLAUDE.md into the rules it gives an agent
3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 55
Document failures (what agents miss without guidance)
- Distinctive: line 56
Write skill addressing specific failures
- Distinctive: line 57
Test with skill (verify improvement)
≈ “Common” is an estimate: at least 1 in 5 of the 26,980 CLAUDE.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
git-repo-prep
Overview
Plugin with two skills for open-source readiness: prepare (full lifecycle) and review (standalone audit). Router in skills/git-repo-prep/SKILL.md dispatches based on user intent.
Skills Included
Main Skill: git-repo-prep
Routes to specific sub-skills based on context:
- Preparing for release →
git-repo-prep:prepare - Reviewing/auditing →
git-repo-prep:review - Ambiguous → AskUserQuestion to clarify
Sub-Skills
- git-repo-prep:prepare — 9-phase interactive workflow (discovery, secrets, license, docs, gitignore, hardening, metadata, backlog, final review). Each phase: scan → present → propose → approve → implement → commit.
- git-repo-prep:review — Category-by-category audit with severity table (Critical/Recommended/Nice-to-have). Conversational output, offers to fix each issue.
When to invoke:
- Making a private repo public
- Preparing a project for open-source release
- Auditing a repo for secrets, missing docs, or openness gaps
- Re-checking after making changes (review mode)
Patterns
Core Philosophy
- Ask, don't prescribe — License choice, CI provider, hardening level all go through AskUserQuestion
- Personal info ≠ secrets — Emails, usernames, and real names are privacy issues separate from credential exposure
- License consistency — Check all three locations: LICENSE file, package metadata, README section
- Openness focus — Stay on topic; no code quality, architecture, or performance review
- Phase-by-phase — Never dump everything at once; work interactively one phase at a time
Ecosystem Awareness
| Indicator | Ecosystem |
package.json | Node.js |
pyproject.toml | Python |
Cargo.toml | Rust |
go.mod | Go |
*.csproj / *.sln | .NET |
Gemfile | Ruby |
pom.xml / build.gradle | Java/Kotlin |
Core checklist is universal. Ecosystem-specific hints (files field, classifiers, engine requirements) added when a known stack is detected.
Development Workflow
Skills were developed using TDD for documentation:
- Baseline tests (pressure scenarios without skill)
- Document failures (what agents miss without guidance)
- Write skill addressing specific failures
- Test with skill (verify improvement)
- Close loopholes (severity drift, missing categories)
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.