product-skills
CLAUDE.md8 starsChanged 9 months agoMITView on GitHub
This repository contains Claude Code skills that are discoverable and installable via skills.sh. These rules are MANDATORY. Violations will cause skills to fail discovery or behave incorrectly. Every SKILL.md file MUST begin with valid YAML frontmatter: Invalid names will cause discovery failures: - ❌ MySkill/ (no camelCase) - ❌ my_skill/ (no underscores) - ❌ skill.md (must be SKILL.md) Your README.md should include these sections: Replace owner/repo-name with your GitHub username and repository name. Include a license section appropriate for your…
In this repository4 other files it gives its agents.
- continuous-discoverySkill
- jobs-to-be-doneSkill
- shape-upSkill
- story-mappingSkill
The file
We split this CLAUDE.md into the rules it gives an agent
5 rules. About 0 are common, 4 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈4
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 11
Every SKILL.md file MUST begin with valid YAML frontmatter:
- Distinctive: line 94
Write as an active collaborator:
- Distinctive: line 99
Create workflows that DO things:
- Distinctive: line 111
Create reference files that just explain concepts
- Distinctive: line 113
Write in passive/academic voice
≈ “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
Skills Repository
This repository contains Claude Code skills that are discoverable and installable via skills.sh.
Critical Requirements
These rules are MANDATORY. Violations will cause skills to fail discovery or behave incorrectly.
YAML Frontmatter (REQUIRED)
Every SKILL.md file MUST begin with valid YAML frontmatter:
---
name: skill-name
description: What this skill does and when to use it. Claude uses this to decide when to apply the skill automatically.
---nameis REQUIRED - must match directory name, lowercase-with-hyphensdescriptionis REQUIRED - Claude uses this for automatic skill selection- Missing frontmatter = skill will not be discovered
Directory Structure (ENFORCED)
your-repo-name/
├── skills/ # ALL skills MUST go here
│ ├── your-skill-name/ # Directory name = skill name
│ │ ├── SKILL.md # REQUIRED - main skill file
│ │ ├── references/ # Optional: supporting documentation
│ │ └── workflows/ # Optional: step-by-step processes
├── CLAUDE.md # Project instructions
└── README.md # Repository documentationFile Naming (ENFORCED)
- Skill directories:
lowercase-with-hyphens - Main file:
SKILL.md(exact casing) - Workflow files:
lowercase-with-hyphens.md - Reference files:
lowercase-with-hyphens.md
Invalid names will cause discovery failures:
- ❌
MySkill/(no camelCase) - ❌
my_skill/(no underscores) - ❌
skill.md(must be SKILL.md)
README Structure
Your README.md should include these sections:
Title and Description
# Your Repository Name
Brief description of what this skills repository provides.Installation
## Installation
\`\`\`bash
npx skills add owner/repo-name
\`\`\`Replace owner/repo-name with your GitHub username and repository name.
Available Skills
## Available Skills
| Skill | Description |
|-------|-------------|
| `/skill-name` | What this skill does |
| `/another-skill` | What this other skill does |License
Include a license section appropriate for your repository.
Skill Design Philosophy
Skills are tools, not documentation. This is not a suggestion—skills that read like reference material fail to provide value.
REQUIRED Patterns
Write as an active collaborator:
- ✅ "I help you find..."
- ✅ "I'll audit this for..."
- ✅ "I check your code for..."
Create workflows that DO things:
- ✅ audit, test, check, analyze, decompose, verify
Focus on application, not explanation:
- ✅ "When you're making a decision, I'll check for these biases..."
- ❌ "Kahneman describes System 1 as..."
PROHIBITED Patterns
DO NOT:
- Summarize what an author said
- Include book-specific stories, examples, or characters
- Create reference files that just explain concepts
- Name skills after books (e.g.,
thinking-fast-and-slow) - Write in passive/academic voice
Example transformation:
Bad (book report):
"Kahneman describes System 1 as fast, automatic thinking..."
Good (thinking tool):
"I'll check this decision for System 1 shortcuts that might be leading you astray."
File Reference Patterns
ALWAYS use relative markdown links. Other patterns will fail.
From SKILL.md to subdirectories
See [workflows/main-workflow.md](workflows/main-workflow.md)
See [references/concepts.md](references/concepts.md)From workflow files to siblings
See [other-workflow.md](other-workflow.md)From workflow files to references (parent directory)
Read [../references/concepts.md](../references/concepts.md)PROHIBITED Link Patterns
These patterns WILL FAIL:
- ❌ Backticks:
`workflows/example.md` - ❌ Plain text:
workflows/example.md - ❌ Absolute paths:
/Users/me/project/skills/... - ❌ No path: just mentioning a filename
Testing & Verification
Local Testing
# Test with Claude Code directly
npx skills add ./path/to/this/repo
# Verify a specific skill loads
# Then run: /skill-nameVerify Discovery Works
# Install from your repo
npx skills add owner/repo-name
# List installed skills
npx skills listCommon Issues
| Problem | Cause | Fix |
| Skill not discovered | Missing/invalid frontmatter | Add required name and description |
| Links don't resolve | Wrong link format | Use [text](relative/path.md) format |
| Name mismatch | Directory ≠ frontmatter name | Make them match exactly |
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.

