meta-skill

Distilly — Distill how they think into reusable Skills for any Agent or Bot. Formerly Colleague Skill（原同事 Skill）.

YAO = Yielding AI Outcomes. A rigorous engineering, evaluation, governance, and portability system for reusable agent skills.

The meta-skill that builds and improves all your skills, including itself. Watches your work sessions (autonomous or human-led), captures patterns, corrections and judgement calls, and turns them into skill improvements and new skill candidates for your review. Practical application of the Augmented Expertise methodology. Open source: CC BY 4.0.

Making ANY Software Skill-Native -- Auto-generate production-ready AI Agent Skills for Claude Code, OpenClaw, Codex, and more.

Scholar-Wendao (学者问道) — open-source Claude Skill framework for distilling scholar perspectives into runnable analytical lenses. Generate perspective skills for humanities scholars (Stiegler, Aristotle, Foucault, …) as research lenses, not chat personas. Bilingual (EN/中文), supports contemporary + traditional scholars. By Shen Cong.

“What’s the thing?” You’d answer: “Skills that teach you how to maintain skills & repos”. Engineering Stewardship for any repository (app, library, tool), a structural governance and quality layer for agent-operated engineering repos.

Claude Code skill for creating new skills with templates and best practices

Your AI agent skill doctor - 5-dimension scoring + security gate + leaderboard

Meta-skill for optimizing Claude Skill descriptions using SDS standard and data-driven analysis

🩺 A meta-skill for Claude Code that gives your other skills a health checkup — diagnose-only, based on Matt Pocock's 'Building Great Agent Skills' method.

pi-agent-harness — a pi coding-agent plugin/package that turns one domain sentence into a multi-agent team: specialist agents (.pi/agents), their skills (.pi/skills), and orchestration prompts (.pi/prompts). A team-architecture factory with 6 patterns and a bundled subagent (single/parallel/chain) delegation tool.

Agent Team & Skill Architect for Claude Code (Simplified Chinese) — 面向 Claude Code 的简体中文元技能，一句话自动生成多智能体团队与技能

Compile any Claude Agent Skill into a multi-agent Workflow — a meta-workflow that recovers the parallelism hidden in a skill's prose and emits a fan-out · pipeline · adversarially-verified Workflow.

Meta-skill that turns any framework's source into a self-contained Agent Skills family — usable later from just the installed package (e.g. in Databricks Genie Code or any AI coding agent).

Health check for Claude Code skills. skill-checkup diagnoses why a SKILL.md won't reliably trigger, walks a dry-run of its workflow, and restructures the package. 4 deterministic scripts, no runtime deps.

WorkBuddy 专家团 Skill 开源包 → DeepSeek Harness (dsh) 插件式 skillgroups 包：44 个多角色专家团队 SKILL.md bundle，支持 ~/.dsh/skills 安装与 Cordis 插件注册。WorkBuddy expert-team skill groups repackaged as a DeepSeek Harness plugin skillgroups pack (44 SKILL.md bundles).

A fork of Anthropic's skill-creator that fixes what the original got wrong: scripts Claude couldn't see, thin trigger tests, and a grader with no path to the tests it was supposed to grade.

⚡ Blazing-fast skill routing engine for AI coding agents. Discovers, indexes, and routes 300+ agent skills in <50ms. Supports Gemini/OpenAI embedding reranking.

baton — author cold-session-executable continuation prompts at session end. 24-section template with macro-context anchor, 9-question self-test, autoresearch loop. Spec-compliant per agentskills.io.

Single self-contained Agentic Loop & Graph Engineering Meta-Skill. Transforms coding agents into Staff-level autonomous engineers. Context Engineering, SDD, 3-Layer Evaluator Loops, Graph G=(V,E), Gauntlet Protocol (Blind Visual Critic), Procedural Memory, and workspace sandboxing.