A collection of AI agent skills for developers and researchers, enhancing productivity and automating complex workflows. It addresses challenges in project planning, multi-LLM strategy generation, documentation access, frontend design, and browser automation. The core benefit is integrating sophisticated AI capabilities for structured planning, rapid information retrieval, and automated browser interactions, streamlining development and research.

How It Works

The project provides modular skills via an npx skills CLI. Key functionalities include Agent Orchestration (planner, plan-harder, llm-council), Documentation Access (context7, openai-docs-skill, read-github), Frontend & Design guidance, and Browser Automation (gemini-computer-use, agent-browser). The llm-council skill uniquely orchestrates multiple LLMs (Claude, Codex, Gemini) to generate and synthesize plans, featuring an interactive, real-time web UI.

Quick Start & Requirements

• Installation: Use npx skills add am-will/codex-skills/skills for all skills, or npx skills add am-will/codex-skills/skills/<skill_name> for specific ones. Global install: npm install -g @am-will/skills.
• Prerequisites:
  • context7: Requires CONTEXT7_API_KEY.
  • gemini-computer-use: Requires GEMINI_API_KEY.
  • llm-council: Requires API keys/subscriptions for Claude, Codex, Gemini; configure via ./setup.sh.
  • Browser tools: Playwright-based (gemini-computer-use) and fast Rust-based (agent-browser, recommended).
• Setup: llm-council includes an auto-launching real-time web UI.

Highlighted Details

• Agent Orchestration: Advanced planning skills (planner, llm-council) for comprehensive, phased implementation plans.
• Multi-LLM Synthesis: llm-council uniquely combines plans from Claude, Codex, and Gemini via a judge agent.
• Browser Automation: Offers Playwright-based and a fast Rust-based headless browser (agent-browser).
• Documentation Access: Skills for querying OpenAI docs, GitHub repos, and general library docs.

Maintenance & Community

The README provides no specific details on maintainers, community channels, or project roadmaps.

Licensing & Compatibility

The README does not specify a license type or compatibility notes for commercial use.

Limitations & Caveats

Several skills require third-party API keys/subscriptions (OpenAI, Google, Anthropic), potentially incurring costs. Users must choose between the two browser automation tools. llm-council requires manual setup via a script.