Jiang-Yude

Jiang-Yude / skill-master

Public

中文版技能包全生命週期管理系統｜整合 Anthropic + OpenAI 官方做法與實戰經驗

10
0
100% credibility
Found Apr 12, 2026 at 10 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

A self-contained skill package that guides users through creating, evaluating, benchmarking, and optimizing custom AI agent skills for environments like Claude and OpenAI.

How It Works

1
📖 Discover Skill Master

You hear about Skill Master, a friendly guide that helps build and perfect custom helpers for your AI chat buddy.

2
🗂️ Add it to your AI

You simply copy the folder into your AI assistant's special skills spot, like Claude or OpenAI, and it's ready to go.

3
💬 Start a conversation

You chat with your AI saying things like 'make a skill' or 'test this skill,' and Skill Master jumps in to help.

4
✨ Build your skill

It walks you through creating your custom skill, from brainstorming ideas to writing clear instructions, step by step.

5
🧪 Run tests and checks

You easily test how well your skill responds to different questions, seeing exactly what works and what doesn't.

6
🔧 Tune for perfection

Skill Master suggests improvements to make your skill trigger at just the right times, boosting its performance.

🎉 Your skill shines

Now you have a reliable, tested custom skill that your AI uses perfectly whenever needed, saving you time and effort.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 10 to 10 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is skill-master?

Skill-master is a Python-based system for managing AI agent skills through their full lifecycle, from ideation to deployment and tuning. It integrates official Anthropic Claude Code skills practices with OpenAI-compatible APIs, letting you create, evaluate, benchmark, and optimize skills via simple voice triggers like "do a skill package" or "run eval." Drop it into Claude Code or OpenAI agents folders, and it handles progressive disclosure, spec-driven design, and cross-environment compatibility without extra installs.

Why is it gaining traction?

It stands out with five workflow modes—guided creation, content conversion, spec-driven builds, automated evals, and trigger tuning—plus scripts for packaging .skill files and generating HTML reports on pass rates. Developers grab it for reliable skill triggering in Anthropic GitHub Claude Code skills or OpenAI setups, avoiding flaky manual tests via forward-testing and benchmark deltas. The Chinese docs and real-world tweaks make it practical for non-English users tweaking Anthropic OpenAI compatible endpoints.

Who should use this?

Prompt engineers building custom skills for Anthropic Claude Code or OpenAI Codex agents, especially those handling computer use tasks or GitHub marketplace tools. Devs iterating on skill-master materia for government-compliant OpenAI and Anthropic setups, or anyone frustrated by low trigger rates in multi-agent flows.

Verdict

With 10 stars and 1.0% credibility score, it's early-stage and lightly tested—docs are solid but expect tweaks for production. Worth forking for Anthropic GitHub skills experiments if you're in the ecosystem.

(187 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.