feiskyer/claude-code-settings
Curated skills, sub-agents, and config templates that supercharge Claude Code — research, image gen, GitHub automation & more.
What it solves
This project provides a comprehensive set of configurations, reusable skills, and specialized sub-agents for Claude Code, extending its native capabilities to handle complex development workflows like deep research, GitHub issue resolution, and design reviews.
How it works
It functions as a plugin or a set of configuration files that can be installed into the Claude Code environment. It introduces "Skills" (reusable capability modules) that Claude can trigger automatically or manually via commands, and "Sub-agents" (specialized personas) for specific roles like UI engineering or PR reviewing. It also includes pre-configured settings.json templates for various LLM providers (e.g., GitHub Copilot, LiteLLM, DeepSeek) to simplify the connection to different model gateways.
Who it’s for
Developers using Claude Code who want to automate repetitive tasks, integrate with external tools (like YouTube transcription or image generation), and implement more structured AI-driven design and research processes.
Highlights
- Diverse Skill Set: Includes tools for brainstorming, deep research (multi-agent orchestration), technical translation, and YouTube transcription.
- GitHub Integration: Specialized skills and agents for fixing GitHub issues end-to-end and reviewing Pull Requests.
- Image Generation: Integrated support for Gemini and OpenAI GPT image models.
- Design Rigor: A "grill-me" skill for high-intensity design reviews and a "handoff" skill for creating session transition documents.
- Multi-Provider Support: Ready-to-use configuration templates for a wide array of model gateways and providers.