# Slite MCP for AI Agents AI Agent Connect > Slite MCP lets you manage your team's knowledge base, notes, and documentation directly through your AI client. It handles everything from searching through nested folders to updating content and verifying docs for accuracy. It is built for teams that need to keep their internal wiki organized without the manual headache of constant clicking. ## Overview - **Category:** knowledge-management - **Price:** Free - **Endpoint:** https://edge.vinkius.com/vk_preview_GS6PfLh90bxJL2dygnNBTC9ZYJTbTW1Esd3oeJhu/ai-agent-connect - **Tags:** documentation, wiki, search-indexing, ai-insights, content-verification, team-collaboration ## Description You spend half your day digging through folders to find a specific internal policy or a project spec that someone updated two weeks ago. This Connector changes that by letting your AI agent act as a librarian for your entire Slite workspace. Instead of clicking through nested hierarchies, you just ask your agent to find the needle in the haystack. It can pull out specific details from deep within your collections or even summarize complex documentation on the fly. It handles the heavy lifting of content management too. If a document starts getting stale, you can have your agent flag it for review or mark it as verified to keep your team's source of truth clean. It is about making sure your team actually uses the knowledge you have already built, rather than letting it sit in a graveyard of unorganized notes. By connecting this to your workflow, you turn a static wiki into a living resource that your AI can actually read, search, and help you maintain. This is where the Vinkius platform makes it easy to bridge your documentation with your agent's capabilities. You will spend less time navigating menus and more time actually making decisions based on the right info. ## Tools ### list_users See a list of all the users in your organization. Use this to manage team access and visibility. ### search_notes Find specific notes using keywords across your entire workspace. This is the fastest way to locate info. ### ask_slite_ai Ask a question to Slite AI to get answers derived from your documentation. This helps you get quick facts without manual searching. ### create_note Create a brand new note in your Slite workspace. Use this to add new docs to your wiki via your agent. ### flag_outdated Mark a specific document as needing a manual review. This keeps your knowledge base from becoming stale. ### get_me Get the profile information for the currently logged in user. This helps the agent know who is making changes. ### get_note Retrieve the full content and details of a specific note. Use this to pull in the full text of a document. ### list_note_children See all the sub-notes nested under a parent note. This helps you understand the structure of your wiki. ### list_collections View all the organized collections in your workspace. This gives the agent a map of your knowledge base. ### list_notes Get a list of every note stored in your Slite workspace. This allows for a full inventory of your documentation. ### update_note Modify the content or details of an existing note. Use this to keep your documentation current. ### verify_note Mark a document as officially verified for your team. This helps users distinguish official info from drafts. ## Prompt Examples **Prompt:** ``` Where can I find the remote work policy in our wiki? ``` **Response:** ``` I found the **Remote Work Policy** in the *Employee Handbook* collection. **Key Details:** * **Eligibility:** All full-time employees. * **Equipment:** Company provides a laptop and monitor. * **Stipend:** $50/month for home internet. You can find the full document in the *Policies* folder. ``` **Prompt:** ``` Summarize the main requirements for the Q4 API project. ``` **Response:** ``` Based on the *Q4 API Project Specs*, here are the main requirements: | Requirement | Status | Priority | | :--- | :--- | :--- | | OAuth 2.0 Integration | Draft | High | | Rate Limiting | Verified | High | | Webhook Support | Planned | Medium | Would you like me to pull up the full documentation for any of these? ``` **Prompt:** ``` Are there any documents that need to be reviewed? ``` **Response:** ``` I've checked the workspace and found **3 documents** flagged as needing review: 1. **Onboarding Guide 2023** (Flagged as outdated) 2. **Security Protocol v1** (Needs verification) 3. **Marketing Assets List** (Flagged as stale) Would you like me to open any of these for you? ``` ## Capabilities ### Search nested note hierarchies Find specific documents buried deep within your workspace folders. ### Create new workspace notes Add new documentation to your knowledge base via your AI client. ### Update existing documentation Modify the content of current notes to keep information current. ### Ask questions to Slite AI Get direct answers derived from your existing documentation. ### Flag outdated content Mark specific documents for review when they become stale. ### Verify document status Mark notes as officially verified for your team. ### List organization users See who is currently part of your Slite workspace. ## Use Cases ### Finding the VPN policy A new hire asks the agent for the VPN setup. The agent uses search_notes to find the correct doc and summarizes the steps. ### Updating the onboarding guide A manager notices a change in the remote work policy and asks the agent to update the guide using update_note. ### Verifying technical specs An engineer wants to know if the API docs are current. The agent checks the status and uses verify_note to confirm. ### Auditing stale content An ops lead asks the agent to find docs that haven't been touched in a year. The agent identifies them for review. ## Benefits - Stop manual folder digging because search_notes finds exactly what you need across your entire workspace instantly. - Keep your team's information accurate by using flag_outdated to catch stale content before it causes confusion. - Speed up content creation by using create_note to build out new documentation directly through your AI client. - Get instant answers on internal policies by using ask_slite_ai to query your existing documentation on the fly. - Maintain a clean workspace by using verify_note to distinguish between official source material and draft content. - Easily organize your workspace by using list_collections to see how your data is structured and nested. ## How It Works The bottom line is you get a way to treat your documentation as a searchable, editable database for your AI. 1. Get your API Key from your Slite workspace settings under the Developer tab. 2. Enter your credentials into the Vinkius platform. 3. Start chatting with your agent to manage your team's knowledge. ## Frequently Asked Questions **How does Slite MCP help my team's documentation?** It allows your AI agent to read, search, and update your Slite workspace. This means your team can get instant answers to internal questions without manual searching. **Can I use Slite MCP to search private notes?** Yes, the agent can search your workspace to find specific notes and details, making it easier to locate information across your team's wiki. **Does Slite MCP help with employee onboarding?** Absolutely. It can quickly pull information from your onboarding guides and documentation to help new hires get up to speed faster. **How do I keep my Slite docs from getting old?** You can use the agent to flag content as outdated or mark it for review. This helps you identify which parts of your knowledge base need attention. **Can my AI agent actually create new pages in Slite?** Yes, the agent can create new notes directly in your workspace. This makes it easy to build out your documentation through a simple chat interface. **Is Slite MCP good for technical documentation?** It is excellent for technical specs and wikis. The agent can navigate nested hierarchies to find specific technical details or summarize complex docs. **Can I ask questions to Slite AI?** Yes, use the ask_slite_ai tool to get answers sourced directly from your documents. **How do I create a nested note?** When using create_note, provide the note_id of the parent in the parent_note_id parameter.