ski

agent
Security Audit
Fail
Health Warn
  • License — License: MIT
  • Description — Repository has a description
  • Active repo — Last push 0 days ago
  • Low visibility — Only 6 GitHub stars
Code Fail
  • rm -rf — Recursive force deletion command in .github/workflows/release.yml
Permissions Pass
  • Permissions — No dangerous permissions requested

No AI report is available for this listing yet.

SUMMARY

🎿 A package manager for agent skills

README.md

🎿 ski

release
checks

ski installs, updates, and links community skills into your coding agent. It scans every file before writing it and records what you installed in ski-lock.json, so your team runs the same reviewed skills.

ski-screen-capture

Features

  • Pick what you install: ski add owner/repo lists the skills in a repository and lets you choose.
  • One install, every agent: Claude Code, OpenCode, and any agent that reads .agents/skills. Each skill is stored once and symlinked into every skills directory.
  • Reviewed updates: ski update compares each installed skill with its source and shows the diff before it changes anything.
  • Dependencies: skills that depend on other skills from the same source are detected, and ski add offers to add them too.
  • Security scan: every file is scanned before it is written. A critical finding stops the install until you approve it. See Security scan.

Install

Homebrew, on macOS or Linux:

brew install osrim/tap/ski

Or download the binary for your platform from the latest release and put it on your PATH. ski needs git. For CI, see Commands.

Quickstart

ski add anthropics/skills   # pick skills, review them, write ski-lock.json
ski install                 # restore every skill in ski-lock.json on a fresh checkout
ski update                  # check upstream and review what changed

Commit ski-lock.json. ski install checks each skill against the integrity recorded in ski-lock.json before writing it, so teammates get the files you approved.

Docs

  • Commands: every command, flag, and exit code, plus non-interactive use.
  • Configuration: scope, agents, paths, environment variables, and the lockfile.
  • Security scan: what the scan catches and what it cannot.

Contributing

See CONTRIBUTING.md.

License

MIT

Reviews (0)

No results found