Slab MCP. Manage your entire knowledge base from chat.

You connect your AI agent straight into your Slab workspace. This isn't just a read-only connection; you give your agent full write access, letting it search, draft, update, and organize documentation across your entire company knowledge base without you ever having to switch apps or copy/paste anything.

What Your Agent Can Do

When you need an answer fast, your agent runs a full-text search against every single article in Slab. You just drop keywords—say, 'Q3 budget approval process'—and the agent pulls back relevant snippets and post IDs from anywhere in the system so you can get answers immediately.

You gotta build out or fix docs sometimes. The agent handles that too. If you need to write a brand new wiki article, it uses create_post; you just give it the Markdown content, and boom—a new post is live in Slab. Need to revise something? You tell it which post ID to hit, and it runs update_post, changing either the title or the body content of that existing article right away.

For structure, your agent manages topics. If a bunch of posts are related—say, all things 'HR Onboarding'—the agent can create an organizing container using create_topic. You don't wanna lose track of stuff, so it also runs list_topics to show you what folders you already got set up. When you need to know the full scope of one folder, the agent uses get_topic_details; this pulls all the metadata for that topic and lists every single post tucked inside it.

Keeping track of everything is key. The agent can list every article ID and its title across your workspace with list_posts. If you just want to know what got changed lately, it hits list_recent_posts for a rundown of the most recently updated documentation so you don't miss a policy change.

When you need specific info on one article, the agent uses get_post_details. You give it a post ID, and it retrieves the full, complete content and all the metadata attached to that single Slab entry. If you wanna see what your whole company setup looks like, the agent pulls the core profile and metadata for your entire organization using get_organization.

For personnel info, the system can run list_users, giving you a complete roster of every registered team member in your Slab workspace. If you need to clean up old content or retire an article, the agent runs archive_post, which permanently removes an existing post from Slab; just remember that action's irreversible through the API.

Basically, it covers the whole lifecycle: finding what exists using search_posts and list_posts, building out structure with create_topic and get_topic_details, writing new stuff with create_post and updating old docs with update_post, pulling user lists via list_users, or just getting the full scoop on a single post using get_post_details. It’s your whole knowledge base, fully accessible through your chat window.