heyabhishekbajpai/skillissue
Developer toolsЭта страница переведена автоматически. Технические детали смотрите в оригинале. Открыть оригинал
Skill Issue — The marketplace for AI skill files. Discover, save, share and combine .md skill files that supercharge your AI agents.
Обзор
Does your AI agent have a skill issue? Probably. Skill Issue is where you find, build, and share .md skill files that make AI agents actually good at specific tasks. Over 10,000 skills indexed from GitHub, searchable in one place. You're a developer dropping skill files into Cursor? A teacher pasting instructions into a Custom GPT? Doesn't matter. Same idea: give your AI a skill file, it stops guessing and starts doing. A .md file with instructions that tell your AI how to behave for a specific task. Instead of explaining everything from scratch every conversation, you drop in a skill and it just knows. No code. No prompt engineering degree required. The LLM reads only the skill name and a short description to decide relevance. The full SKILL.md (often thousands of tokens) loads only when it actually matches. Token efficient by design.
README
Показан фрагмент оригинального README. Открыть оригинал.
skill issue.
Does your AI agent have a skill issue? Probably.
Skill Issue is where you find, build, and share .md skill files that make AI agents actually good at specific tasks. Over 10,000 skills indexed from GitHub, searchable in one place.
You’re a developer dropping skill files into Cursor? A teacher pasting instructions into a Custom GPT? Doesn’t matter. Same idea: give your AI a skill file, it stops guessing and starts doing.
Live at skillissue.bajpai.tech
what even is a skill file?
A .md file with instructions that tell your AI how to behave for a specific task. Instead of explaining everything from scratch every conversation, you drop in a skill and it just knows.
No code. No prompt engineering degree required.
The LLM reads only the skill name and a short description to decide relevance. The full SKILL.md (often thousands of tokens) loads only when it actually matches. Token efficient by design.
what you can do
- Browse 10,000+ skills from Anthropic, Vercel, OpenAI, HuggingFace, and hundreds of community repos
- Build your own skill by describing what you want in plain English. The AI writes the SKILL.md for you
- Download as
.zipor copy straight to clipboard - Save privately to your vault or publish for everyone
- Combine multiple skills into a single package
- Share any skill with a direct link
who actually uses this
A lawyer who wants their AI to draft contracts without hallucinating case law. A teacher who needs better student notices. A YouTuber tired of scripts that sound like a robot wrote them. A developer whose coding agent keeps making the same mistakes.
All of them end up here for the same reason: they want their AI to be better at one specific thing, and a skill file does that.
architecture
The system has two separate data pipelines. The crawler finds skills on GitHub and stores metadata in MongoDB. The frontend fetches actual skill content directly from GitHub when you open one. Content is never stored, only metadata.
How the crawler works:
A cron job runs every 15 days. It hits GitHub’s Code Search API with 6 size partitions (up to 6000 results), filters repos through GraphQL batches (50 per call, minimum 10 stars), then deep scans each repo with the Trees API to find every SKILL.md file. Deduplicates by repo::path, bulk upserts into MongoDB in batches of 500. The whole run takes about 5-8 minutes.
the stack
Рекомендуемые инструменты
Попробуйте другой запрос или уберите фильтр.
Установка
npx skillfish add heyabhishekbajpai/skillissue