Frequently asked questions
Everything you might want to know about Skills Manager, agent skills, MCP, access control, security, and pricing.
What is Skills Manager?
It is a cloud control plane that centralizes your organization's AI agent skills, controls who can use which skill (org, team, or person), and serves them securely to Claude and any MCP client — synced from Git.
What are agent skills?
Agent skills are reusable units of agent know-how — workflows, SOPs, scripts and instructions, defined by the open standard at agentskills.io — that an AI agent can load and run. The hub stores, governs, and serves them.
What is MCP?
MCP (Model Context Protocol) is the de-facto open standard for connecting AI clients to tools and context. It is native in Anthropic, OpenAI, Google and Microsoft products, has tens of millions of monthly SDK downloads, and was donated to the Linux Foundation's Agentic AI Foundation in December 2025.
Which AI clients are supported?
Any MCP-compatible client. That includes claude.ai, Claude Desktop, Claude Code, and Codex. Desktop apps use the OAuth 2.1 connector; CLI tools use scoped personal access tokens.
How does access control work?
You grant each skill org-wide, to a specific team, or to an individual user. You can additionally scope a token or connection to a subset of skills — for example, give one agent only three skills.
How do I connect my AI client?
For claude.ai / Claude Desktop, add a custom MCP connector pointed at https://www.skillsmanager.uk/mcp and authorize via OAuth 2.1. For Claude Code / Codex, generate a scoped personal access token and add it to your MCP config. See the How it works page for examples.
Where do my skills actually live?
In your own Git repositories — Git is the source of truth. The hub syncs from your GitHub repos and publishes a curated view. There is no lock-in.
Can users create their own skills?
Yes. Users can author personal skills directly in the hub, and every change is automatically pushed to their own Git repository as a backup.
How are changes to shared skills governed?
Feedback and edit proposals route to admins for review. Approved changes are committed back to Git, giving you an auditable approval trail for every modification.
Is it secure?
Yes. It uses Supabase Auth with row-level org isolation, Vault-encrypted GitHub keys, OAuth 2.1, scoped tokens, and least-privilege access. See the Security page for details.
How is my organization's data isolated from others?
Through database row-level security (RLS). Every query is constrained to the caller's organization, so tenants cannot access each other's data.
Can I manage everything from my AI client?
Yes. Alongside the web admin UI, the hub exposes 40 MCP tools so you can add sources, grant access, mint tokens, and more directly from Claude.
What are favorites?
Skills you mark as favorites are preloaded into the agent's context, so your most-used playbooks are immediately available each session.
How much does it cost?
There is a Free tier, a Team plan for shared access-controlled libraries, and an Enterprise plan with SSO and isolation. Pricing on the pricing page is indicative — contact us for current figures.
How do I get started?
Sign up, connect a Git source, grant access, and connect your AI client. You can be running governed skills within minutes.