skill-library
CLAUDE.md7 starsChanged 8 months agoMITView on GitHub
A Claude Code marketplace containing 41 agent skills organized into 6 plugin families, extracted from 109 PDFs using Google Gemini. Each plugin bundles skills that encode domain expertise and process discipline. This is a content repository — no build system, no tests, no application code. This repo is a marketplace — install the whole thing or individual family plugins. Each family is a standalone plugin with its own .claude-plugin/plugin.json. Skills within a plugin are independently invocable. For standalone use without…
The file
We split this CLAUDE.md into the rules it gives an agent
2 rules. About 0 are common, 1 is distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
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
- Distinctive: line 92
Source Attribution: Every principle and framework cites its source (book or URL).
- Very specific: line 94
Writing New Skills: Follow meta:writing-skills — TDD approach.
≈ “Common” is an estimate: at least 1 in 5 of the 27,641 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
CLAUDE.md
This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
What This Is
A Claude Code marketplace containing 41 agent skills organized into 6 plugin families, extracted from 109 PDFs using Google Gemini. Each plugin bundles skills that encode domain expertise and process discipline. This is a content repository — no build system, no tests, no application code.
Repository Structure
.claude-plugin/
└── marketplace.json # Marketplace definition (lists 6 plugins)
plugins/
├── meta/ # 6 skills — FULL rigor (process enforcement)
│ ├── .claude-plugin/plugin.json
│ ├── skills/{name}/SKILL.md
│ └── README.md
├── workflow/ # 12 skills — FULL rigor (strategic discovery-to-delivery)
├── executive/ # 9 skills — STANDARD rigor (leadership & org execution)
├── narrative/ # 4 skills — STANDARD rigor (fiction & storytelling)
├── rhetorician/ # 5 skills — STANDARD rigor (non-fiction writing)
└── dealmaker/ # 5 skills — STANDARD rigor (negotiation & interpersonal)GEMINI.md— Master specification: architecture, templates, resource mappings, extraction methodology (120KB)PROGRESS.md— Batch completion report (41/41 skills, project finalized)
Plugin System
This repo is a marketplace — install the whole thing or individual family plugins.
# Install all plugins via marketplace
claude install /path/to/skill-library
# Skills are then invocable as plugin-name:skill-name
# e.g., meta:getting-started, workflow:problem-framing, dealmaker:negotiation-tacticianEach family is a standalone plugin with its own .claude-plugin/plugin.json. Skills within a plugin are independently invocable.
For standalone use without the plugin system, copy any skills/{name}/SKILL.md directory into a project's .claude/skills/ folder.
Two Rigor Levels
Full rigor (meta + workflow families, 18 skills): Enforce process. Each has:
- Iron Law — A single non-negotiable constraint (e.g., "NO SKILL WITHOUT A FAILING TEST FIRST")
- State Machine — Graphviz
dotdiagram defining workflow states and transitions - Rationalization Table — Maps common shortcut-thinking to reality
- Red Flags — Thoughts that indicate process drift
- Diagnostic Checklist — Testable acceptance criteria
Standard rigor (executive, narrative, rhetorician, dealmaker families, 23 skills): Encode expertise. Each has:
- Guiding Principles — Foundational ideas with source citations
- Frameworks & Models — Domain-specific mental models
- Common Mistakes — Typical failure modes
- Diagnostic Checklist — Validation criteria
Skill File Format
Each skill lives at plugins/{family}/skills/{skill-name}/SKILL.md with YAML frontmatter conforming to the agentskills.io specification:
---
name: kebab-case-name
description: "What the skill does + when to activate it (under 1024 chars)"
license: MIT
compatibility: "Claude Code and compatible agent products"
metadata:
type: "meta | workflow | domain"
family: "meta | workflow | executive | narrative | rhetorician | dealmaker"
rigor: "full | standard"
keywords: "comma, separated, search, terms"
requires: "skill-a, skill-b" # Prerequisite skills (comma-separated)
enhances: "skill-c, skill-d" # Skills improved by combination
sources_pdf: "Book Title (Author)" # Book citations (comma-separated)
sources_web: "Source Name" # Web citations (comma-separated)
---Top-level fields (name, description, license, compatibility) are spec-recognized. All other fields live under metadata as string values. Claude Code reads name and description for routing; the remaining fields are metadata for tooling and human reference.
Key Architectural Concepts
The 1% Rule (meta:using-skills): If there is even a 1% chance a skill applies, invoke it before responding. This is the core enforcement mechanism.
Session Bootstrap (meta:getting-started): Routes new tasks to the correct skill. Start here.
Skill Chaining: Skills declare requires and enhances dependencies in metadata (as comma-separated strings). The workflow family forms a chain: problem-framing → stakeholder-discovery → market-context → competitive-analysis → buyer-persona → business-case → deliverable skills (prd-writing, pitch-deck, one-pager, executive-briefing) → assumption-audit → stakeholder-review.
Source Attribution: Every principle and framework cites its source (book or URL). No ungrounded methodologies.
Writing New Skills: Follow meta:writing-skills — TDD approach. Define the user problem, write a failing test case, then author the skill content. New skills go in plugins/{family}/skills/{skill-name}/SKILL.md.
Conventions
- Skills live in
plugins/{family}/skills/{skill-name}/SKILL.md - Skill names and directories:
kebab-case - Each family plugin has a
README.mdlisting its skills - State machines use Graphviz
dotsyntax (not Mermaid) - Iron laws are backtick-wrapped ALL-CAPS sentences
- Source citations inline:
(Source: Author, *Title*) - Plugin metadata:
.claude-plugin/plugin.jsonper family,.claude-plugin/marketplace.jsonat root
Writing Voice
When drafting non-fiction prose, state things directly. Do not use contrast framing ("This isn't X, it's Y"), engagement bait ("But here's the thing," "Here's what most people get wrong"), or contrarian hooks ("The real question is…"). If the point is strong, it does not need a theatrical frame.
Linear
Linear project: Dev Tooling (team: Joellewis)
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.

