Connect an AI assistant to the docs
The docs run an MCP server at https://docs.unitelabs.io/mcp. Connect your AI assistant to it and the assistant can look up guides, method signatures and class members itself, instead of guessing from what it learned in training. It needs no account and no API key.
Add the server
claude mcp add --transport http unitelabs-docs https://docs.unitelabs.io/mcp
Open the Cursor install link, or add the server to .cursor/mcp.json in your project:
{
"mcpServers": {
"unitelabs-docs": {
"url": "https://docs.unitelabs.io/mcp"
}
}
}
Open the VS Code install link, or add the server to .vscode/mcp.json in your project:
{
"servers": {
"unitelabs-docs": {
"type": "http",
"url": "https://docs.unitelabs.io/mcp"
}
}
}
In Settings → Connectors, choose Add custom connector, name it UniteLabs Docs and enter https://docs.unitelabs.io/mcp as the URL.
Any other client that speaks MCP over streamable HTTP works with the same URL.
Try it
Ask a question that needs the reference, for example:
Using the UniteLabs docs: which method measures liquid heights on a Hamilton STAR,
and what parameters does it take?
The assistant should call search or get-symbol, find Channels.get_liquid_heights, and answer with its signature and parameters.
What the assistant can call
| Tool | What it returns |
|---|---|
search | Ranked guide sections and API symbols for a few keywords, each with a link. |
get-symbol | One class, function or method: signature, parameters, return value, raised exceptions, examples, docstring, and the guides whose code uses it. |
list-members | Every public method of a class, own and inherited, plus its attributes. Start here for questions like "what can a STAR do?" |
list-pages | All guide pages with title, description and type. |
get-page | The full markdown of one page. |
The API reference covers the latest release of the Python SDK, the Liquid Handling SDK, Labware, the Omnibus and the CDK. For an older version, open its entry in the library's changelog, for example the Liquid Handling SDK changelog.
Without MCP
If your tool cannot connect to an MCP server, point it at plain text instead:
https://docs.unitelabs.io/llms.txtlists every page with a link to its markdown, plus one API digest per library.- An API digest, such as
https://docs.unitelabs.io/reference-digest/python-liquid-handling-sdk.md, lists every public module, class and function of the latest release, one line each, with a link to its full entry. https://docs.unitelabs.io/llms-full.txtis every guide page in one file.- Copy page at the top of every page copies that page as markdown.
Last updated