Claude Skills Collection
A collection of custom skills and agents for Claude Code that extend AI capabilities with specialized workflows, domain knowledge, and best practices.
📘 Read the full story - Learn about our philosophy, architecture, and roadmap.
Available Skills
Python coding standards and patterns for user-friendly and maintainer-friendly code.
Friendly Python provides comprehensive guidelines for writing Pythonic code that's both easy to use and easy to maintain. Based on Frost Ming's "Friendly Python" series, this skill encapsulates best practices for API design, code organization, and common patterns.
- Registry Patterns - Replace if-else chains with extensible registries
- Context Managers - Proper resource management with
with statements
- Classmethod Constructors - Multiple input sources with clean APIs
- Descriptors - Explicit field definitions over
__getattr__ magic
- Extension Points - Use official hooks instead of custom wrappers
- Python Idioms - async/await, decorators, generators over callbacks
Key Features:
- 🎯 7 proven code patterns with good/bad examples
- ✅ Review checklist for code quality assessment
- 📚 Quick reference tables for common scenarios
- 🐍 Pythonic over ported patterns from other languages
- 🔍 Type hints and discoverability first
Use when:
- Writing new Python code
- Reviewing and refactoring existing code
- Designing public APIs
- Porting code from other languages to Python
- Training team members on Python best practices
→ Learn more about friendly-python
Maintains project context through progressive disclosure and enforced documentation.
Context Keeper implements a three-level documentation system that ensures AI agents always have accurate, up-to-date context about your codebase:
- USERAGENTS.md - Project-level structure, conventions, and directory index
- TECH_INFO.md - Directory-level file descriptions and dependencies
- File headers - File-level function descriptions and parameters
Key Features:
- 🔍 Automatic tech stack detection (TypeScript, React, Go, Python, etc.)
- 📋 Coding conventions inference based on detected stack
- 🗂️ Smart directory filtering using
.gitignore
- 🔄 Optional stop-time documentation checks via hooks
- 📝 Enforced documentation updates via AGENTS.md instructions
Use when:
- Starting work on a new project
- Onboarding new team members (or AI agents)
- Maintaining context across long development sessions
- Working with complex codebases
→ Learn more about context-keeper
Automate Chrome via Puppeteer to browse, navigate, scrape and capture web pages.
Auto-Browser provides comprehensive tools for web automation and interaction:
- Browser Control - Launch Chrome instances and manage browser lifecycle
- Page Navigation - Open URLs, handle redirects, and manage page state
- DOM Interaction - Interactive element picking and JavaScript execution
- Content Capture - Take screenshots, extract HTML, and scrape web content
Key Features:
- 🌐 Full Chrome/Chromium automation via Puppeteer
- 🖱️ Interactive DOM element selection tools
- 📸 Screenshot and visual capture capabilities
- 💻 Execute custom JavaScript on pages
- 📄 HTML extraction and web scraping
- 🔄 Session management and cookie handling
Use when:
- Automating web interactions and testing
- Scraping content from websites
- Taking screenshots of web pages
- Testing web applications
- Gathering information from dynamic websites
→ Learn more about auto-browser
Product Requirements Manager for ideation and product management.
Ideate-PM helps manage product design information by turning a single product idea into an evolving requirements collection. It maintains a requirements database with associated design decisions and open questions:
- Requirements Management - Add, refine, and track product requirements with IDs
- Decision Tracking - Record design decisions and their rationale
- Open Questions - Track unresolved questions that need answers
- State Management - Query current product design state and backlog
- Markdown Views - Generate human-readable views from database
Key Features:
- 📋 SQLite-based authoritative storage for requirements
- 🎯 Single product per repository management
- 📝 Markdown view generation (PRODUCT.md, BACKLOG.md, OPEN_QUESTIONS.md)
- 🔍 Requirement status tracking (PROPOSED, READY, DONE)
- ⚖️ Priority management (P0, P1, P2)
- 🗂️ Decision and question tracking with scope awareness
