RoleCraft

Install AI agent skills as roles & behaviors — from any source.
Zero-dependency CLI. No marketplace. No signup.

Works with 65+ AI agents: opencode · claude-code · cursor · windsurf · devin · codex · copilot · aider · cline · gemini-cli · cody · continue · warp · codeium · fabric · goose · tabnine · supermaven · pr-pilot · loom · roo · trae · hermes · kiro · augment · kilo · openhands · junie · factory · command-code · cortex · mistral-vibe · qwen-code · openclaw · codebuddy · mux · pi · autohand-code · rovo · firebender · bob · aider-desk · and more

Quick Start · Why RoleCraft? · Commands · Install Guide · Contribute

Quick start

# try without installing
npx rolecraft --help

# or install globally (works with npm, pnpm, yarn, bun)
npm install -g rolecraft

# create a skill
rolecraft init my-skill

# install it
rolecraft install ./my-skill                      # local folder
rolecraft install user/repo                       # GitHub repo
rolecraft install https://gitlab.com/org/project  # GitLab repo
rolecraft install git@github.com:user/repo.git    # SSH URL
rolecraft install ./my-skill --cursor             # specific agent only

# manage
rolecraft list
rolecraft search code-review
rolecraft check
rolecraft remove my-skill

Requirements: Node.js >= 20 · No other dependencies · 65+ agents supported · Full install guide →

---

Features

• Zero dependencies — ~4 KB, no bloat
• Any source — local folder, GitHub/GitLab/Bitbucket repo, SSH git URL
• 65+ agents — opencode, claude-code, cursor, copilot, aider, devin, gemini-cli, and more
• No registry required — no signup, no marketplace, no vendor lock-in
• Non-interactive mode — --yes / -y flag for automation/CI pipelines
• Update checking — rolecraft check to see which skills have updates
• Shell completions — bash, zsh, fish auto-completion
• TUI search — interactive arrow-key skill browser with preview
• Content hash verification — detect tampered or outdated skills
• CI-ready — lockfile-based re-install for pipelines
• Dry-run mode — preview before installing

---

Commands overview

Command	Description	Details
rolecraft init [<name>]	Scaffold a new SKILL.md	docs
rolecraft install <source>	Install a skill (local path, GitHub/GitLab/SSH URL)	docs
rolecraft bundle <sources>	Install multiple skills from inline sources or file	docs
rolecraft bundle create	Create a new bundle file	docs
rolecraft search <query>	Search for skills on GitHub (TUI with --interactive)	docs
rolecraft check	Check installed skills for available updates	docs
rolecraft use <source>	Preview a skill's files without installing	docs
rolecraft completions bash|zsh|fish	Generate shell completion scripts	docs
rolecraft setup [<source>]	Detect agents, optionally install a skill to all	docs
rolecraft list	Show all installed skills	docs
rolecraft verify	Check installed skill integrity via content hash	docs
rolecraft ci	Re-install all skills from lockfile (CI mode)	docs
rolecraft upgrade	Upgrade rolecraft to the latest version	docs
rolecraft remove <slug>	Uninstall a skill	docs
rolecraft update <slug>	Re-install a skill to latest	docs
rolecraft --version	Show version

---

Why rolecraft?

→ Full feature comparison

Feature	rolecraft	skills (Vercel)	@agentskill.sh/cli
Zero dependencies	✅	✅ (1 dep)	❌ (2)
Local path install	✅ 1st class	✅	❌ marketplace only
GitHub repo install	✅	✅	❌
GitLab / SSH git URL	✅	✅	❌
Agent targets	66	55+	15+
Bundle install + create	✅	❌	✅ (skillset only)
Interactive TUI search + install	✅	✅	❌
Non-interactive flag (--yes/-y)	✅	✅	❌
Skill update check (check)	✅	❌	❌
Shell completions (bash/zsh/fish)	✅	❌	❌
Dry-run preview (--dry-run)	✅	❌	❌
Interactive scope prompt	✅	✅	❌
Content hash verification (verify)	✅	✅	❌
CI-mode re-install (ci)	✅	✅	❌
Self-upgrade command	✅	❌	❌
File size	~4 KB	~465 KB	~84 KB

See full table →

---

How agents discover skills

rolecraft knows where each AI agent looks for skills. Use flags like --claude, --cursor, --devin to target specific agents, or --all for every supported agent.

→ Full agent path table

# Install to multiple agents at once
rolecraft install ./my-skill --cursor --devin --copilot --gemini --cody

---

Architecture

1. Reads SKILL.md from the source and parses metadata (slug, name, owner)
2. Copies (or symlinks with --symlink) all files alongside SKILL.md to the target directory
3. Computes a SHA256 content hash and stores it in the lockfile
4. Updates ~/.agents/.skill-lock.json so agents can discover the skill
5. Compatible with skills installed by @agentskill.sh/cli, add-skill, or manual installs

→ Full architecture & project structure

---

Contributing

Contributions are welcome! See CONTRIBUTING.md for guidelines.

License

MIT