craigspaterson/agent-skills
Reusable skill definitions that teach AI agents consistent output formats, writing guidelines, and worked examples — works with Claude Code, GitHub Copilot, Cursor, and more.
Reusable skill definitions that teach AI agents consistent output formats, writing guidelines, and worked examples — works with Claude Code, GitHub Copilot, Cursor, and more.
npx add-skill craigspaterson/agent-skillsReusable skill definitions that teach AI agents consistent output formats, writing guidelines, and worked examples — works with Claude Code, GitHub Copilot, Cursor, and more.
Local Pokemon Red LLM-agent harness with live gameplay UI, structured traces, save states, replay, and turn-by-turn agent observability.
An opinionated, Dockerized OpenCode environment with a built-in AI engineering team — agents, skills, and slash commands for any tech stack.
MCP server that lets Claude Code and other AI agents read large PDFs without hitting context limits. Chunked reading, hybrid search, OCR, table and image extraction, SQLite cache.
AI chat application featuring branching message trees, real-time streaming responses, multi-format artifact rendering (Markdown, HTML, XLSX, Mermaid diagrams), file handling, and integrated Model Context Protocol (MCP) servers
Code, Build and Evaluate agents - excellent Model and Skills/MCP/ACP Support
Hand off a coding requirement and watch AI agents (Claude Code, Codex, Gemini CLI) ship merged PRs. Decomposes the work, runs stories in parallel git worktrees with per-story ephemeral Postgres, reviews and tests every change, escalates failures through 5 tiers, merges what passes. Event-sourced, Go, Mac/Linux.