agentleFS
iberi22iberi22/skills/AGENTS.md

skills

AGENTS.md0 starsChanged 36 days agoMITView on GitHub

This repository is a curated registry of agent skills — reusable expertise modules that can be loaded by AI agents (e.g., via SWAL, Codex, Claude Code, or any agent framework that supports skill injection). Each skill lives in skills/<category>/<skill-name>/ and is defined by a SKILL.md file with YAML frontmatter. Reference a skill by its id (declared in frontmatter). Example: You can also load all skills within a category: Each skill exposes a raw_url pointing to its SKILL.md on GitHub (raw).…

The file

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

3 rules. About 0 are common, 0 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Create a folder: skills/<category>/<skill-name>/

    line 125
  2. Very specific:

    Add SKILL.md with valid YAML frontmatter and markdown body.

    line 126
  3. Very specific:

    Run python3 scripts/manifest-generator.py to regenerate the manifest.

    line 128

≈ “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

AGENTS.md — Agent Briefing for iberi22/skills

Overview

This repository is a curated registry of agent skills — reusable expertise modules that can be loaded by AI agents (e.g., via SWAL, Codex, Claude Code, or any agent framework that supports skill injection).

Each skill lives in skills/<category>/<skill-name>/ and is defined by a SKILL.md file with YAML frontmatter.


How to Use These Skills

1. Load by ID

Reference a skill by its id (declared in frontmatter). Example:

skills:
  - id: nextjs
  - id: python
  - id: content-creator

2. Load by Category

You can also load all skills within a category:

skills:
  - category: marketing

3. Raw URL Access

Each skill exposes a raw_url pointing to its SKILL.md on GitHub (raw). This is useful for dynamic loading at runtime:

https://raw.githubusercontent.com/iberi22/skills/main/skills/<category>/<skill-name>/SKILL.md

4. Manifest

The canonical registry is at _registry/manifest.yaml. It contains metadata for all skills: id, name, description, category, tags, path, raw_url, authors, etc. Parse this to discover available skills programmatically.


Format Conventions (v2)

All SKILL.md files follow the v2 format:

---
id: unique-skill-id
name: Display Name
category: category-name
tags:
  - tag1
  - tag2
goals:
  - "One-line description of what this skill enables."
  - "Another goal or context."
authors:
  - Author Name
---

# Content

Markdown body with system prompt, instructions, examples, and constraints.

Required Frontmatter Fields

FieldTypeDescription
idstringUnique kebab-case identifier
namestringHuman-readable name
categorystringTop-level category folder name
tagsstring[]Searchable tags
goalsstring[]What the skill enables (1–3 items ideal)

Optional Frontmatter Fields

FieldTypeDescription
versionstringSemantic version
descriptionstringShort description (auto-extracted if missing)
authorsstring[]Attribution
licensestringLicense identifier
original_sourcestringUpstream URL if ported
contextstringWhen/where to apply this skill

Category Taxonomy

CategoryDescriptionCount
aiAI/ML tooling, fine-tuning, inference—
communicationWriting, messaging, tone, translation—
debuggingBug triage, root-cause analysis, diagnostics—
designUI/UX, visual storytelling, brand, whimsy—
devopsDeployment, infrastructure, CI/CD—
educationLearning content, exam generation, validation—
engineering-principlesArchitecture, code quality, principles—
engineering-workflowTDD, prototyping, rapid iteration, testing—
frameworkNext.js, Astro, Vite, Tailwind CSS, etc.—
frontendFrontend-specific agents and patterns—
languageLanguage-specific guidelines (Python, Rust, etc.)—
marketingContent, growth, social media, ASO—
operationsAnalytics, finance, legal, support—
productResearch, feedback, prioritization—
project-managementSprints, shipping, studio coordination—
testingQA, benchmarking, workflow optimization—
toolsCLI tools, agent wrappers, MCP integrations—
webWeb design guidelines, accessibility, performance—

Note: Category counts are auto-generated in _registry/manifest.yaml.


Contributing

Adding a New Skill

  1. Create a folder: skills/<category>/<skill-name>/
  2. Add SKILL.md with valid YAML frontmatter and markdown body.
  3. Optionally add companion files (e.g., setup.md, examples/).
  4. Run python3 scripts/manifest-generator.py to regenerate the manifest.
  5. Open a Pull Request.

PR Checklist

  • SKILL.md frontmatter is valid YAML (run a YAML linter).
  • id is unique across the entire registry.
  • category matches an existing folder or is added to taxonomy.
  • goals are concise and actionable.
  • No sensitive data (API keys, passwords, internal URLs) in content.

CI / Lint

Pull requests are validated by GitHub Actions:

  • YAML lint — all SKILL.md frontmatters must parse cleanly.
  • ID uniqueness — no duplicate skill IDs.
  • Manifest freshness — manifest.yaml must be up to date (or the manifest workflow will auto-commit it after merge).

Automation

WorkflowTriggerPurpose
manifest.ymlPush to skills/**Regenerates _registry/manifest.yaml and commits it.
skill-watchdog.ymlWeekly (cron) + manual dispatchMonitors upstream repos for new skills and opens issues.

Questions?

Open an issue or reach out via the repo discussions.

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.