agentleFS
Sign inStart free
alton47alton47/threejs-skills/AGENTS.md

threejs-skills

AGENTS.md13 starsChanged 7 months agoMITView on GitHub

This file is for AI coding agents (Claude, Copilot, Cursor, Codex, etc.) that have write access to this repository. This repo contains AI Skills for Three.js — structured markdown files that teach AI agents how to use Three.js correctly. The goal is concise, accurate, non-hallucinated Three.js knowledge. Before committing skill changes, test with: Open an issue or PR if you spot a hallucinated API, outdated pattern, or missing skill domain.

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

11 rules. About 0 are common, 3 are distinctive, and 8 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.

  • 8

    Very specific

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

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

    Do not hallucinate APIs: If you are unsure whether a method exists in Three.js, say so or omit it.

    line 11
  2. Very specific:

    Every skill folder must contain exactly one SKILL.md

    line 27
  3. Very specific:

    All reference files linked from SKILL.md must exist in the same folder

    line 29
  4. Very specific:

    No nested references (don't link from SHADOWS.md → CASCADES.md — link everything from SKILL.md)

    line 30
  5. Distinctive:

    Do not add marketing language or fluff to skills

    line 34
  6. Very specific:

    Do not over-explain what Three.js is — Claude already knows

    line 35
  7. Very specific:

    Do not add redundant comments in code examples (// This creates a box above new THREE.BoxGeometry())

    line 36
  8. Very specific:

    Do not copy-paste full Three.js source code into skills — reference and explain patterns instead

    line 37
  9. Distinctive:

    Before committing skill changes, test with:

    line 41
  10. Very specific:

    Ask a representative task (see examples/ for reference)

    line 44
  11. Distinctive:

    Verify Claude's output is correct, uses modern APIs, and doesn't hallucinate

    line 45

≈ “Common” is an estimate: at least 1 in 5 of the 36,368 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 — Guidance for AI Agents Editing This Repo

This file is for AI coding agents (Claude, Copilot, Cursor, Codex, etc.) that have write access to this repository.

Purpose of This Repo

This repo contains AI Skills for Three.js — structured markdown files that teach AI agents how to use Three.js correctly. The goal is concise, accurate, non-hallucinated Three.js knowledge.

When Editing Skills

  • Do not hallucinate APIs. If you are unsure whether a method exists in Three.js, say so or omit it. Do not invent method names, parameter names, or class names.
  • Pin to Three.js r160+ unless explicitly noted. The Three.js API changes frequently. Always note the version if you reference something that changed.
  • Keep SKILL.md under 500 lines. Move detailed reference to separate .md files in the same folder and link to them.
  • Use forward slashes in all file paths (scripts/helper.py, not scripts\helper.py).
  • Prefer one recommended approach over listing 5 equivalent options. Include an escape hatch if a secondary approach is commonly needed.
  • Do not add time-sensitive information like "as of March 2024" — this will become stale.

Naming Conventions

  • Skill folders: threejs-{domain} (lowercase, hyphens only)
  • Reference files: UPPERCASE.md (e.g., SHADOWS.md, GLSL.md)
  • Example files: kebab-case.js or kebab-case.jsx
  • Scripts: snake_case.py or camelCase.js

Structure Rules

  • Every skill folder must contain exactly one SKILL.md
  • SKILL.md must have valid YAML frontmatter with name and description
  • All reference files linked from SKILL.md must exist in the same folder
  • No nested references (don't link from SHADOWS.md → CASCADES.md — link everything from SKILL.md)

What Not to Do

  • Do not add marketing language or fluff to skills
  • Do not over-explain what Three.js is — Claude already knows
  • Do not add redundant comments in code examples (// This creates a box above new THREE.BoxGeometry())
  • Do not copy-paste full Three.js source code into skills — reference and explain patterns instead

Testing

Before committing skill changes, test with:

  1. A fresh Claude Code session with the skill loaded
  2. Ask a representative task (see examples/ for reference)
  3. Verify Claude's output is correct, uses modern APIs, and doesn't hallucinate

Contact

Open an issue or PR if you spot a hallucinated API, outdated pattern, or missing skill domain.

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