GUIDE
How to turn a spreadsheet into an MCP server
Updated
Export your sheet as a CSV and upload it to Kaiva Bridge. Bridge checks the columns with you, imports the rows and builds tools for them, then hosts them as an MCP server your AI can query. The data is read-only for the AI.
Step by step
- Export a CSVFrom Excel, Google Sheets or Numbers, export the sheet as CSV.
- Create a spreadsheet serverChoose New server, then Spreadsheet, and upload the CSV.
- Check the columnsBridge shows each column with the type it detected (text, number, date). Correct anything it guessed wrong, then import.
- Connect your AI clientWhen the import is done the server is live. Copy the address from the Connect tab and add it to your AI client.
Good to know
You can save a filtered view as its own tool, for example "orders shipped last week", so your AI calls it by name.
To refresh the data, upload a new CSV. Free includes 3 MCP servers, 200,000 tool calls a month and 200 pages of documents, with no card.
Other ways to do it
- Load it into a databaseImport the CSV into PostgreSQL or MySQL and connect the database instead. Better for data that changes often.
- Attach the file to one chatFine for a single question; a spreadsheet server works across every client and conversation.
Questions
- Which formats work?
- CSV. Export from Excel, Google Sheets or Numbers first.
- Can the AI change my data?
- No. Spreadsheet servers only read.
- How do I update the data?
- Upload a new CSV to the same server.
- Is it free?
- Free includes 3 MCP servers, 200,000 tool calls a month and 200 pages of documents, with no card.