fang2houfang2hou/skills/AGENTS.md

skills

AGENTS.md2 starsChanged 3 months agoMITView on GitHub

This repository is a collection of Agent Skills for coding agents that need reusable, task-focused workflows. This repository has no test, build, or dev mise tasks. mise run check verifies Markdown formatting with oxfmt, runs the root skill-structure check, and runs the redacted Gitleaks scan; mise run format rewrites Markdown in place. Follow the shared engineering guidelines in the AI Coding Guidelines portal. Read the portal recipe for the task before editing; repository documentation is the source of truth when…

Where files are headingNew AGENTS.md files have more rules about keeping secrets out of the code, and fewer about code style.
Of the AGENTS.md files we read, the 31,542 first seen in the last six months, against the older ones: the kind of rule that gained most share, and the kind that lost most. About files like this in general, not this one, and an estimate, since a kind is a pattern in the words.
In this repository10 other files it gives its agents.
- agent-reportSkill
- create-skillSkill
- figcraft-liveSkill
- git-branch-cleanupSkill
- github-repo-baselineSkill
- herdr-parallel-devSkill
- improve-skillSkill
- natural-writingSkill
- review-and-fixSkill
- test-designSkill

The file

Show its rules and text

We split this AGENTS.md into the rules it gives an agent

14 rules. About 0 are common, 3 are distinctive, and 11 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.
11
Very specific
Tied to their project's names, files or tools. Probably skip.

Common
Distinctive
Very specific

Distinctive:
Follow the shared engineering guidelines in the AI Coding Guidelines portal.
line 17
Very specific:
Run mise run check and require exit status 0 before every commit.
line 33
Very specific:
When changing a file under skills/, keep the skill entrypoint at skills/*/SKILL.md and run bash scripts/check-skills.sh before opening a …
line 34
Distinctive:
When adding a relative link in a root document, point it to an existing path in this repository and recheck the link before committing.
line 35
Very specific:
Add package.json, pnpm-lock.yaml, tsconfig.json, or JavaScript/TypeScript source files.
line 39
Very specific:
Commit a change when mise run check exits nonzero.
line 40
Very specific:
Add another AGENTS.md below skills/.
line 41
Very specific:
Before adding a top-level directory or changing mise.toml, .oxfmtrc.json, .pre-commit-config.yaml, .gitattributes, .github/workflows/, or …
line 46
Very specific:
Before changing the install or update commands in README.md, including npx skills add fang2hou/skills -g or npx skills update -g -y.
line 47
Very specific:
Before deleting or renaming any skills/*/SKILL.md or changing a published skill's frontmatter name.
line 48
Distinctive:
Do not infer UI language from conversation language.
line 59
Very specific:
Keep each skill's references/ directory one level deep.
line 64
Very specific:
Write every skill in English.
line 65
Very specific:
After changing a skill, use npx skills add fang2hou/skills -g to verify the installed distribution on this machine.
line 67

≈ “Common” is an estimate: at least 1 in 5 of the 40,246 AGENTS.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.

Keep what fits and drop the rest. Your version, ready for your agents.

Files

AGENTS.md

This repository is a collection of Agent Skills for coding agents that need reusable, task-focused workflows.

Commands

mise install
mise run format
mise run check

This repository has no test, build, or dev mise tasks. mise run check verifies Markdown formatting with oxfmt, runs the root skill-structure check, and runs the redacted Gitleaks scan; mise run format rewrites Markdown in place.

Engineering Standards

Follow the shared engineering guidelines in the AI Coding Guidelines portal. Read the portal recipe for the task before editing; repository documentation is the source of truth when it is more specific.

Project-specific overrides:

None. This repository follows the shared standards.

Layout

.github/ — GitHub Actions workflows and the pull request template.
scripts/ — repository validation scripts, including check-skills.sh.
skills/ — the ten installable Agent Skills; each has a SKILL.md and may have one-level references/, scripts/, or assets/ content.

Boundaries

Always:

Run mise run check and require exit status 0 before every commit.
When changing a file under skills/, keep the skill entrypoint at skills/*/SKILL.md and run bash scripts/check-skills.sh before opening a pull request.
When adding a relative link in a root document, point it to an existing path in this repository and recheck the link before committing.

Never:

Add package.json, pnpm-lock.yaml, tsconfig.json, or JavaScript/TypeScript source files; this repository validates Markdown and Bash instead.
Commit a change when mise run check exits nonzero.
Add another AGENTS.md below skills/; the root file is the canonical agent instruction surface.
Change skills/herdr-parallel-dev/scripts/verify-worktrees.sh during a skill-content change unless the request names that path.

Ask first:

Before adding a top-level directory or changing mise.toml, .oxfmtrc.json, .pre-commit-config.yaml, .gitattributes, .github/workflows/, or scripts/check-skills.sh.
Before changing the install or update commands in README.md, including npx skills add fang2hou/skills -g or npx skills update -g -y.
Before deleting or renaming any skills/*/SKILL.md or changing a published skill's frontmatter name.

Confirmed Language Policy

Item	Value
Conversation	Follows the language used by the user
Code / comments / commits	English
UI language	No UI in this repository; skill content follows the target document language
Tone	Clear, concise, and technical; use the target document's natural register

Do not infer UI language from conversation language.

Project Conventions

A skill directory's name must exactly match the name field in its SKILL.md frontmatter.
Keep each skill's references/ directory one level deep; keep SKILL.md below 500 lines when the workflow can remain complete and clear.
Write every skill in English. The frontmatter, When to Use, and When NOT to Use contain no Chinese, Japanese, or Korean text; non-English text appears only as an example inside a rule about that language or as UI strings a skill ships.
scripts/check-skills.sh validates frontmatter delimiters, the lowercase skill-name format, the exact directory/name match, and a non-empty description no longer than 1024 characters; it warns above 250 description characters or 500 body lines.
After changing a skill, use npx skills add fang2hou/skills -g to verify the installed distribution on this machine; on another machine, refresh it with npx skills update -g -y.

Depth: DEVELOPMENT.md for workflow and toolchain, CONTRIBUTING.md for pull request rules.

Download now (ZIP)

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

What did you use it for? e.g. Reviewing pull requests in a Rails app

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.

Check your setup, free