Work with AI Assistants (MCP)
DocStar has an MCP server. MCP (Model Context Protocol) is the standard way AI assistants such as Claude, ChatGPT, Cursor, VS Code and Codex connect to other apps. There are two things you can do with it.
Build an MCP server for your own app
Have an API? Hand it to your AI assistant and ask it to turn it into an MCP server. The assistant reads your latest code, documents each endpoint in a DocStar collection, and DocStar gives that collection its own MCP server. Your users add one URL to Claude, ChatGPT, Cursor or VS Code, then use your app in plain language.
You stay in control of how it behaves:
- Tools. Pick up to 20 of your most important endpoints as tools of their own.
- Skill. Write the guide assistants read before using your API.
- Instructions. Add rules assistants follow every time they connect.
- Sign-in. Let users call your API with their own account.
See Build an MCP server for your app.
Do your documentation work with AI
Connect DocStar to your AI assistant and ask for documentation work in plain language.
Write a "Getting Started" page in our Help Docs collection and publish it.
Read our Express routes and document every endpoint in the Payments API collection.
Fix the typos on the Webhooks page, but show me the changes before you save.
Your assistant finds the right workspace, collection and page, writes the content, and asks before it changes anything.
How it works
- You add the DocStar MCP URL to your AI assistant once.
- The first time the assistant needs your workspace, you sign in to DocStar in your browser. There is no API key to copy.
- The assistant works only with the workspaces and collections your DocStar account can access, with the same permissions you have.
If your assistant can't connect, you can use an Auth Key instead. See Run the DocStar API with an Auth Key.
What you can do
| You want to | Ask your assistant to |
|---|---|
| Build an MCP server | Read your API code, document the endpoints, and set up the tools, skill and instructions. |
| Write docs | Create a page, or a set of pages, in a collection. |
| Update docs | Read a page, rewrite or fix it, and save the new version. |
| Document an API | Read your code and create endpoint docs with requests and responses. |
| Publish | Publish pages and endpoints when you're ready. |
| Organise | Find pages, list collections, rename pages and change their URLs. |
| Give feedback | Send a bug report or feature request straight to the DocStar team. |
Safe by design
- It asks first. The assistant shows you what it will create, change or publish, and waits for your OK.
- Changes start as drafts. Every edit is saved as a draft. Nothing reaches your live docs until it is published.
- Your permissions apply. If you can only read a collection, so can your assistant.
- Mistakes can be undone. Deleted pages and collections can be restored for 90 days, and every page keeps its version history.