Most teams have done the work. They’ve built a knowledge base with accurate, well-maintained documentation. They’ve adopted AI assistants to move faster. But the two tools have never spoken to each other.
The AI doesn’t know what’s in the knowledge base. The knowledge base gets nothing from the AI. And every time someone needs both, they become the bridge, pulling content from one place, pasting it into another, waiting for an output, then repeating the whole cycle. Two powerful tools sitting on the same team, never actually working together.
Document360 is a dedicated knowledge base software built for support, documentation, and product teams. With its newly launched MCP server support, AI assistants like Claude and ChatGPT get a live, direct connection to that content, no manual handoff required.
This post covers what MCP is, what the Document360 integration enables, which teams benefit most, and how to get started.
The Problem With Disconnected AI and Documentation
Most AI assistants are useful general-purpose tools. They can draft, summarise, reformat, and answer questions at speed. But they only know what they have been trained on, or what someone pastes into them in the moment. Without a direct connection to your knowledge base, they are working blind.
This creates a familiar workaround. Someone pulls an article from the knowledge base, drops it into a prompt, gets a response, and pastes the result somewhere useful. It works for a single interaction. It does not work at scale. And every time the knowledge base is updated, the cycle starts again from scratch.
The deeper problem is consistency. When AI tools cannot access your actual documentation, the answers they generate start to drift from what is officially written. A support agent gets an AI-generated response that contradicts the help center. An internal wiki grows content that does not match the published docs. The knowledge base and the AI pull in different directions without anyone noticing until something goes wrong.
MCP addresses this at the connection layer. Rather than manually bridging the gap each time, Document360’s MCP server makes your knowledge base a live resource that AI assistants can query, retrieve from, and write back to, all within your existing governance model.
Document360 MCP: What the Feature Actually Enables
MCP stands for Model Context Protocol. It is an open standard that gives AI assistants structured, secure access to external systems. Document360 has built its own MCP server, so any MCP-compatible AI tool can now connect to your knowledge base without custom development or third-party middleware.
What AI Assistants Can Do
Through the Document360 MCP server, AI assistants can search for articles across project versions, retrieve full article content by article ID or URL, create new articles, and update existing content. These operations cover the full cycle of documentation work, from research to publishing, all accessible within an AI conversation.
Security Follows Your Existing Rules
The integration is built on OAuth. When an AI assistant connects to Document360 through MCP, the connection is tied to the user account that completes the OAuth flow. All interactions follow Document360’s role-based access control model. The AI can only access content that the authenticated user is permitted to see. Restricted articles, unpublished drafts, and content with limited visibility all stay governed by the same rules already in place. Write operations go through your existing workflow and versioning rules, not around them.
Supported AI Tools
Supported AI assistants include Claude, ChatGPT, GitHub Copilot, and Cursor IDE. Any platform that supports MCP or custom connectors can also be connected.
For teams with specific integration requirements, Document360 supports a custom connector path that extends compatibility beyond the default list.
This is a feature built into your existing Document360 project. There is no infrastructure to provision and no separate service to manage. It works within your current project configuration and governance setup from the moment it is enabled.
Check out this video on how to connect any AI tool to a knowledge base using MCP- https://www.youtube.com/watch?v=V60PGzefWm0
How Different Teams Use Document360 MCP
The same MCP server powers different workflows depending on the team. Teams do not need separate configurations. MCP surfaces content based on each user’s permissions and project settings.
Support Teams
When a ticket arrives, support engineers can ask their AI assistant to search Document360 for relevant articles based on the ticket context. The response is grounded in actual product documentation, not AI inference. If no article exists, they can draft one directly from the same session and push it back into the knowledge base, closing the gap without switching tools.
Technical Writers and Content Teams
Instead of manually searching for what already exists before starting a new article, writers can ask their AI assistant to retrieve related content, identify coverage gaps, and draft new material with real context from the knowledge base. Research and writing happen in a single session. The output is grounded in what the team has already documented.
Product Managers
When preparing for a release, product managers can use MCP to pull documentation by project version, compare how features are described across releases, and flag inconsistencies before they reach customers. Documentation review becomes part of the pre-launch process rather than something that gets picked up after the fact.
Developers
Developer teams using AI-assisted IDEs like Cursor can connect Document360 through MCP to retrieve version-specific documentation without leaving their coding environment. Draft documentation for new features or endpoints can also be generated directly in the IDE, while the context from the development work is still fresh.
Enabling MCP in Document360
MCP is available in Document360 now. Enabling it requires a Project Owner or Admin role. Connecting to an AI assistant uses a standard OAuth flow and does not require engineering resources for most teams.
Once connected, Document360 provides a dedicated MCP prompt guide with ready-made prompt patterns to help teams get practical value from the integration quickly. There is also published guidance on when to use MCP versus Document360’s standard API, a useful reference for teams that want to understand where each interface fits best. MCP is built for AI assistant interactions. The API is better suited for programmatic, app-to-app integrations. They are designed for different use cases and work well alongside each other.
Enabling MCP is one thing. Knowing whether your team is getting value from it is another. Document360 includes an MCP server analytics dashboard that provides visibility into how AI assistants and tools interact with your knowledge base. It tracks total calls, success and failure rates, and how many unique users are making MCP requests within a given period. There are breakdowns by tool type so you can tell whether usage leans toward content retrieval or content creation, plus a chart showing which AI assistants are connecting most frequently. Search activity is tracked separately from other tool calls. One practical note: the dashboard stays accessible even after MCP is disabled, so past activity remains reviewable for teams that need to report on usage retroactively.
For teams already using Document360, MCP is an additive capability. Nothing about how you currently manage your knowledge base changes. It extends what your AI assistants can do without touching your existing setup.
One Connection, Less Friction
The separation between AI tools and knowledge bases has been one of those workflow gaps that teams route around without expecting it to be fixed. Document360’s MCP server closes that gap with a direct, governed, live connection between the AI assistants your team already uses and the documentation you have already built.
Whether the priority is faster support resolution, more consistent documentation output, or better developer workflows, MCP makes the knowledge base an active part of how work gets done. Not a separate system people remember to check. A live resource that is always part of the conversation.
If your team is already on Document360, the feature is available now. The MCP prompt guide in the docs is a practical place to start.
