Docs / Connect Mindola
Ways to connect
Use Mindola from Claude, ChatGPT, your code editor, Slack, or your own scripts, not only from the app.
You do not have to be inside the Mindola app to build or grow a lens. A lens is your public AI page at one shareable link, and Mindola runs a hosted MCP server plus a small REST API so other tools can create lenses, add knowledge, and check progress for you.
The MCP server is remote. There is nothing to download or run locally: you add one URL to your client and your assistant gains three tools. One creates a lens from a profile you approve in the conversation, one adds knowledge to it, and one reports ingestion status. No account or API key is needed to start. Anonymous use creates an unclaimed lens that is live right away, with a claim link so you can make it yours later. The full tool reference lives in MCP server.
AI assistants
- Claude: add the Mindola connector in claude.ai, Claude Desktop, or Claude Code, then ask Claude to create your lens in conversation. See Claude.
- ChatGPT: add the Mindola connector in developer mode. ChatGPT uses its own endpoint,
https://app.mindola.ai/mcp. See ChatGPT.
Code editors and CLIs
Any editor or CLI that supports remote MCP servers can connect with the same URL, https://app.mindola.ai/mcp/lens. Each page below shows the exact config snippet:
- Cursor: a small block in
mcp.json. See Cursor. - VS Code: a
.vscode/mcp.jsonentry. See VS Code. - Codex CLI: a few lines in
config.toml. See Codex. - Windsurf: an entry in its MCP config. See Windsurf.
- Gemini CLI: an entry in
settings.json. See Gemini CLI.
Using something else? Any client that speaks Streamable HTTP works. See Other clients.
Slack
Connect a lens to a Slack workspace so people can mention it in a channel and get answers from its scoped knowledge, with the same privacy rules as the web link. See Slack.
Scripts and your own code
The REST API lets you create lenses, add knowledge, and poll status from your own code with an API key. See the API reference and the runnable examples in curl, Node, and Python. The MCP documentation and examples are also open source at github.com/Mindola-ai/mcp-server.
Where to start
If you just want to see what a lens feels like, try the live demo at mindola.ai/me/charles-darwin. If you already chat with Claude or ChatGPT, the connector there is the fastest way to a lens of your own. And the app at app.mindola.ai always gives you the full picture: knowledge, appearance, visitors, and settings in one place.