my-claude-agent-skill
AI pair programming
Documentation
My Claude Agent Skills
A collection of custom skills for Claude Code Agent to enhance AI-assisted development workflows.
Overview
This repository contains reusable skills that extend Claude's capabilities for specific development tasks. Each skill provides structured workflows and best practices for common engineering scenarios.
Skills
1. AI Pair Programming
Location: `ai-pair-programming/SKILL.md`
A flexible dual-AI engineering workflow that supports custom AI pairing (Claude, Codex, Gemini, GPT-4, etc.) for collaborative development.
Key Features:
- Customizable AI pairing: Choose your preferred Primary AI (for planning/implementation) and Review AI (for validation/review)
- Balanced AI collaboration: Primary AI handles architecture, planning, and execution while Review AI provides validation and code review
- Continuous feedback loop: Each AI reviews the other's work for optimal code quality
- Structured phases: Planning → Validation → Execution → Cross-Review → Iterative Improvement
Supported AI Configurations:
| Configuration | Primary AI | Review AI |
|---|---|---|
| Claude + Codex | Claude | Codex |
| Claude + Gemini | Claude | Gemini |
| Gemini + Claude | Gemini | Claude |
| GPT-4 + Codex | GPT-4 | Codex |
Workflow:
Plan ({{PRIMARY_AI}}) → Validate Plan ({{REVIEW_AI}}) → Feedback →
Implement ({{PRIMARY_AI}}) → Review Code ({{REVIEW_AI}}) →
Fix Issues ({{PRIMARY_AI}}) → Re-validate ({{REVIEW_AI}}) → Repeat until perfect2. Figma to Jetpack Compose
Location: `figma-to-compose/SKILL.md`
Automates the generation of Android Jetpack Compose UI code from Figma designs using MCP (Model Context Protocol) tools.
Key Features:
- Automatic UI generation: Converts Figma frames/components to idiomatic Jetpack Compose code (Material3)
- Design token extraction: Extracts colors, typography, spacing, and radius from Figma variables
- Icon pipeline: Automatically detects icon/vector nodes and converts them to Android VectorDrawable XML
- Debug helpers: Includes crash/ANR debugging utilities using ADB
Required MCP Tools:
- Figma Desktop MCP: `get_metadata`, `get_variable_defs`, `create_design_system_rules`, `get_design_context`, `get_screenshot`
- Android MCP Toolkit: `convert-svg-to-android-drawable`, `read-adb-logcat`, `fetch-crash-stacktrace`, `check-anr-state`
Output Deliverables:
- Screen composables with previews
- Reusable component files
- Theme files (Color.kt, Type.kt, Dimens.kt)
- VectorDrawable XML icons
Usage
These skills are designed to be used with Claude Code Agent. To use a skill:
1. Reference the skill in your Claude Code configuration or conversation
2. Provide the required inputs as specified in each skill's documentation
3. Follow the structured workflow phases
Contributing
Feel free to submit issues or pull requests to improve existing skills or add new ones.
License
This project is available for personal and educational use.
Frequently asked questions
What is my-claude-agent-skill?
my-claude-agent-skill is AI pair programming
How do I install my-claude-agent-skill?
Open the GitHub repository and follow its README. Most MCP servers are added to your client's MCP config, then called by your agent.
Is my-claude-agent-skill open source?
Yes — it is hosted on GitHub at https://github.com/Nam0101/my-claude-agent-skill and has 2 stars.
Related MCP tools
Model Context Protocol Servers
Pre-indexed code knowledge graph, auto syncs on code changes, for Claude Code, Codex, Gemini, Cursor, OpenCode, AntiGravity, Kiro, CoPilot, and Hermes Agent — fewer tokens, fewer tool calls, 100% local
an open source, extensible AI agent that goes beyond code suggestions - install, execute, edit, and test with any LLM
The easy-to-use open source Business Intelligence and Embedded Analytics tool that lets everyone work with data :bar_chart:
The Open-Source Multimodal AI Agent Stack: Connecting Cutting-Edge AI Models and Agent Infra
Cognee is the open-source AI memory platform for agents. Give your AI agents persistent long-term memory across sessions with a self-hosted knowledge graph engine.
Run your own MCP server? See who uses it and what to fix.
Measure it with TrackMCP