SkillRepo is the open distribution and governance layer for AI agent skills.
One graded library. Each repo gets the skills it needs, at the versions you approve — and any drift in plain sight.
One layer for distribution and governance, three ways in.
Whether you run a team, publish skills, or build the agents that will use them — here's where each path leads.
Standardize agent skills across your team
One curated, graded library — scoped to each repo at the versions you approve — instead of five private setups drifting apart.
Publish agent skills and get discovered
A permanent home, GitHub attribution, and reach analytics — so your best work isn’t a link in a year-old blog post.
Where agents are headed
Every operation an agent runs is becoming a skill it loads — where the whole system is heading next.
From scattered files to shared infrastructure.
Packages got npm. Containers got Docker Hub. Python libraries got PyPI.
Skills get SkillRepo.
A format emerged. Distribution didn’t.
The standard showed up — agentskills.io, adopted by 30+ tools. The distribution layer didn’t. A format without distribution is packages without npm. Vendor CLIs handle the install. The registry is just the catalog; the distribution layer handles what to install, from whom, graded how, and what your team curates — delivered to every agent you run, and kept current.
Not the same set everywhere — the right set in each repo.
Distribution gets the library to every agent. Governance decides what each repo runs: give each repo the skills it needs, declared in git; pin the version you approved; see which repos have drifted. When a bad version gets through, recall it across your organization’s repos.
Know what you’re pulling — and who wrote it.
Every skill is attributed to the GitHub identity we found— sign up with GitHub to claim yours, and your own skills stay private by default. 36.8% of public skills contain a security flaw (Snyk, Feb 2026), so every publish gets two-layer analysis: static checks plus an AI pass across safety, quality, clarity, completeness. Each skill ships with an A–F grade and the findings behind it. Grades are signals, not a certification. Always review a skill yourself before use.
Two ways this changes your week.
Your skills, discoverable.
- A permanent URL and namespace tied to your GitHub identity — claimed once, yours.
- Grading and install counts show which of your skills land.
- Your publisher page is the portfolio — every skill, every install, one link to share.
Five engineers. Five agents. Five slightly different versions of the same skill.
You don't realize it until the output diverges — your codebase starts looking like it was written by five different teams because every agent was working from a different copy of “how we do things.”
A team library on SkillRepo flips it. You curate one library; each repo gets the skills it needs, at the versions you approve — and you see which repos have drifted. Your code starts looking like your team wrote it.
Teams from $8/seat/month. No seat minimum. 14-day trial, no card.
The short version.
- What is skill distribution?
- Publish once, and the current approved skill files land where your agents already read them — and stay current.
- What is skill governance?
- Skill governance is control over who approved a skill, which version each repo runs, what has drifted, and what to recall.
- What is SkillRepo?
- SkillRepo is the open distribution and governance layer for AI agent skills — one graded library your team can’t drift from, delivered to every agent.