# Fygment: The platform for managing and governing AI skills.

Bring your skills together, keep track of every version, and use them in the AI tools you already work with: Claude, ChatGPT, Gemini, Copilot, Cursor and more.

- [Get started](https://fygment.dev/signup) (free, self-serve)
- [Documentation](https://fygment.dev/docs)
- [Agent guide (llms.txt)](https://fygment.dev/llms.txt)
- [MCP server setup](https://fygment.dev/docs/mcp/setup)
- [REST API (OpenAPI)](https://fygment.dev/openapi.json)
- [REST API reference: versioning, rate limits, errors](https://fygment.dev/docs/api)

External skills get a security check with [NVIDIA SkillSpector](https://github.com/NVIDIA/SkillSpector) before anyone uses them: brought in, checked and approved, then the same version delivered to your tools.

## See how it works for you

Fygment works inside the AI tools your team already uses. A connected assistant such as Claude finds the approved team skill for a task, uses it, and reports back. Fygment keeps track of which version was used and whether it is current.

## Know what a skill does before you let it in.

Every external skill, for example one from a public GitHub repository, is scanned before it is imported. You see what it does and then decide: import it, ask for changes, or block it.

## Full control over how your skills are governed.

Set each rule for the whole company, a team or a single skill.

### Who can create and share

Decide who may create or import skills, and how widely each person can share them: with themselves, their team, or the whole company.

### Where skills can come from

Allow only the sources you trust, for example your company GitHub and never public sources.

### Which AI tools can use them

Choose the AI tools that may receive your skills, for example Claude and Copilot but not personal accounts.

### What needs approval, and by whom

Choose what must be reviewed before anyone can use it, and who gives the approval: a skill owner, an admin, the security team, or two approvers for critical skills.

### How updates reach people

Offer the latest reviewed version or pin one, decide how strongly updates are pushed (optional, recommended or required), and roll out to one team first.

### How often skills are re-reviewed

Give each skill a risk level. The riskier it is, the more often its owner reviews it again: critical every 30 days, high every 90 days, medium every 6 months, low every year.

## Free to start. Pay when your team needs control.

Start with a local skill inventory for free. Upgrade when imports, AI review, team governance and enterprise controls become part of how you work.

### Solo

For one person. **Free forever.** A simple local setup before you bring in external sources or team workflows.

- Local skill inventory
- Create and package skills
- Install and check in supported AI tools
- Public templates and examples

### Advanced / Teams

For teams and AI imports. **$14 per seat per month.** AI-assisted skill intake, a private registry and governed rollout for your team.

- AI-assisted imports from GitHub, ZIP and npm
- Version control and update checks for imported skills
- Security and compatibility review
- Private team skill registry
- Roles and approvals
- Audit trail and install drift alerts

### Enterprise

For regulated and large organisations. **Custom pricing.** Identity, deployment, compliance and audit depth for organisations with strict requirements.

- Everything in Advanced / Teams
- SSO (SAML, OIDC) and SCIM
- Policy-as-code with PR review
- On-prem registry and air-gapped sync
- Customer-managed signing keys
- SOC 2 Type II and ISO 27001 support
- Unlimited audit retention and SIEM export

## Bring your first skill in today.

Create a workspace, connect the AI tool you already use, and put a skill to work.

- [Get started](https://fygment.dev/signup)
- Prefer the terminal? Run `npx fygment` ([CLI docs](https://fygment.dev/docs/cli)).
- [Support](https://fygment.dev/support) · [Privacy](https://fygment.dev/privacy) · [Terms](https://fygment.dev/terms)