Skills

AI agent skills published by Andrea Griffiths

• Agent Work Contracts
  Turn multi-step or delegated agent work into a bounded contract with concrete validation and a concise handoff. One skill that works across Copilot, Claude Code, Cursor, Codex, Gemini CLI, and OpenClaw via the Agent Skills standard.
  Agent Skills Agent Plugins Handoff Repository ↗
• Dream Cycle
  Nightly AI research scanner that surfaces relevant improvements and generates proposals. Scans arxiv papers, GitHub trending repos, filters by relevance, and generates scored improvement proposals.
  Agent orchestration RAG Voice agents AI safety
• GitHub Actions
  Debug failing workflows, write maintainable CI/CD pipelines, and avoid common GitHub Actions pitfalls. Covers gh CLI debugging, local testing with act, composite actions, SHA pinning, caching, and concurrency groups.
  gh CLI act Dependabot
• Context & Prompt Management
  Automatically manage token limits and optimize prompts for LLM workflows. Accumulate context from multiple sources, intelligently summarize when limits are exceeded, and generate prompts that fit within model constraints.
  GitHub Models API Python
• Video Subtitles
  Generate accurate subtitles for videos using AI transcription. Supports multiple languages, timestamps, and export formats.
  Whisper FFmpeg
• Last 30 Days Copilot
  Track and analyze GitHub Copilot usage and metrics over the last 30 days. Generates reports on completions, acceptance rates, and productivity impact.
  GitHub Copilot
• Publish to GitHub Pages
  Automate deployment of static sites to GitHub Pages. Handles build, optimization, and deployment workflows.
  GitHub Pages GitHub Actions Repository ↗
• Copilot Cost Audit
  Audit any repo for GitHub Copilot cost optimizations. Checks copilot-instructions.md for output controls, scans for MCP server overhead, flags hardcoded premium model names, and reviews agent code for prompt caching.
  GitHub Copilot CLI