# Nuclino MCP > Nuclino MCP gives your AI agent full read/write access to corporate knowledge bases, team structures, and project documentation. It treats your entire wiki—including nested workspaces, collections, and files—as one searchable source. You can query company history, draft new specs, or track user roles directly inside your preferred client. ## Overview - **Category:** productivity - **Price:** Free - **Tags:** team-wiki, document-collaboration, workspace-management, knowledge-base, project-tracking ## Description Connect your AI agent to your internal knowledge base for reading, writing, and organizing corporate data. This MCP lets you interact with your company's entire workspace structure—everything from team listings down through individual documents. You can ask the agent to find a specific policy across multiple departments or draft an update directly into a living project page. It tracks documentation evolution over time, making sure any new guideline or meeting note gets recorded instantly. Through Vinkius, you get one connection point that lets your AI client access this deep organizational data without needing dozens of separate API keys. ## Tools ### list_teams Lists all organizational teams the authenticated user belongs to, helping locate specific departments within your company structure. ### list_workspaces Returns a list of isolated workspaces mapped within a specified team, allowing you to scope subsequent queries accurately. ### list_items Lists all standard knowledge items (pages) in a workspace, giving you an overview of the available documentation titles. ### get_item Retrieves the precise Markdown payload and full configuration details for a specific document item. ### create_item Writes a brand new knowledge page or wiki item into a specified workspace, making it immediately visible to your team. ### update_item Appends new edits or overwrites partial content inside an existing documented item. ### delete_item Permanently removes a structural knowledge item from the wiki after confirming with you. ### search_items Executes a global semantic search across an entire team's content to find unknown documents or policies by keyword. ### list_collections Lists the document collections that segment and organize a target workspace, helping you understand the document relationship graph paths. ### list_fields Maps customizable structured property fields available globally across an entire team's knowledge base. ### list_users Lists all human identities attached to a specific team, giving you a roster of active members and owners. ### list_files Retrieves the URLs for physical attachments bolted onto an item, allowing you to access source files like PDFs or images. ## Prompt Examples **Prompt:** ``` Search Nuclino for any documentation mentioning 'SSO Security Policies'. ``` **Response:** ``` I've successfully performed a fuzzy text search and located the page 'SSO Security Implementation Guidelines'. I can extract its contents or share the browser link if needed. ``` **Prompt:** ``` Create an item titled 'Project X Architecture Brief' in the Engineering workspace. ``` **Response:** ``` I retrieved the Engineering workspace ID and created the new item 'Project X Architecture Brief'. The document is live and awaiting further edits. ``` **Prompt:** ``` List all teams connected to this authentication token. ``` **Response:** ``` You have access to 3 teams on this environment: Design Ops (`id-1`), Software Engineering (`id-2`), and Executive Planning (`id-3`). ``` ## Capabilities ### Map the entire team structure Discover every root team and nested workspace ID within your organization's hierarchy. ### Find specific documents globally Run a fuzzy search across all knowledge items to pinpoint exact policies, notes, or project pages no matter where they live. ### Read and analyze content Retrieve the full Markdown payload of any item, along with listing all associated physical files and attachments. ### Draft new documentation Instantly create a brand new knowledge page or document within an existing workspace using natural language prompts. ### Manage structure and users View who is attached to the team, map customizable fields, or list all collections that group related documents. ## Use Cases ### Onboarding a new team member A new PM needs to understand the department's process flow. Instead of asking three different people for links, they ask their agent to search all documentation mentioning 'onboarding workflow'. The agent runs `search_items` and compiles the top five relevant pages. ### Updating a critical policy The compliance team needs to revise an old security guideline. They tell their agent, which uses `get_item` to pull the existing Markdown content. The agent drafts the revision and then executes `update_item`, ensuring the change is live immediately. ### Finding a specific attachment The engineer needs the final PDF spec sheet for Project Chimera, but they only know it was attached to 'Project Alpha'. They use the agent to find the item and list all files using `list_files` to get the direct download link. ### Structuring a new project A department is launching a massive initiative. The PM asks their agent to create the main planning hub, which uses `create_item`. They then ask the agent to list all available team members using `list_users` to set up initial roles. ## Benefits - Stop digging through nested folders. Use `search_items` to run a global query across your entire team's wiki, finding policies or specs regardless of which workspace they live in. - Drafting updates used to mean copying old text into a new doc. Now, you can use the agent to read existing content via `get_item`, draft the changes, and then update it using `update_item`—all in one go. - You don't have to manually track who owns what. The MCP lets you list all team members (`list_users`) and map out ownership fields (`list_fields`), keeping your documentation accurate automatically. - Need a new planning document? Instead of creating it manually, the agent can use `create_item` instantly, generating a live, structured page that awaits content input. - The MCP exposes structural details like collections and workspaces. You get visibility into how your knowledge base is organized via tools like `list_collections`, which saves hours of mapping time. ## How It Works The bottom line is that your AI agent treats your corporate wiki like a database, giving you structure and content access all in one go. 1. Subscribe to this MCP and enter your Nuclino Personal API Key. 2. Select a task in your AI client (Claude, Cursor, etc.), telling the agent what knowledge you need or want to create. 3. The agent uses its tools—like `search_items` or `create_item`—to interact with your wiki and deliver the result. ## Frequently Asked Questions **How does Nuclino MCP search for documents?** The agent performs indexed semantic searches across an entire team using `search_items`. This means it doesn't just match keywords; it understands the context of your query to find related content. **Can I use Nuclino MCP to update a document?** Yes, you can. The agent uses `update_item` to append new text or overwrite partial states in an existing wiki page, keeping your documentation current without manual copy-pasting. **Does Nuclino MCP only list items by title?** No. Beyond listing titles with `list_items`, you can read the exact content and configuration of any item using `get_item` to see the full Markdown payload, including hidden metadata. **What if I need to find a file attached to an item?** You use the `list_files` tool. It exposes pure URL bindings for physical attachments on any specific document item, letting you access source files like PDFs or images. **How do I start using Nuclino MCP in my AI client?** You subscribe to this MCP on Vinkius and provide your Nuclino Personal API Key. After that, you just tell your agent what documentation task needs completing.