Ask your AI what is actually in your warehouse.
The datasets in your BigQuery project, the tables and views inside each one, and every column’s name and type — answered inside ChatGPT, Claude, Codex or Cursor, so the person writing the query starts from the real schema. Connect BigQuery to toolfare once; your AI reads it with your own credentials, and every read is free.
1Connect your AI app
Pick the app you use. You do this once, and every API on toolfare is then available in it.
Add toolfare as a connector in Claude
Add to Claudehttps://api.toolfare.ai/mcp
- 1Open Customize → Connectors and choose "Add custom connector".
- 2Paste this URL as the remote MCP server URL.
- 3Leave the OAuth client id and secret under Advanced settings empty.
- 4Complete the browser authorization, then return to your conversation.
Custom connectors are available on Free, Pro, Max, Team and Enterprise. A Free account may hold one custom connector in total.
Claude’s own instructions2What to ask
Copy a prompt into your AI app once you’re connected. It picks the right one of the 4 tools and tells you the price first.
The datasets in a project
Every dataset the connection can see, with its region — the starting point for knowing what data exists at all.
- “What datasets do we have in BigQuery?”
- “Which of our datasets are in the EU region?”
The tables and views inside a dataset
Tables, views, materialised views and external tables in one dataset, each with its type — views are often the curated entry point, and they are listed alongside tables.
- “What tables are in our marketing dataset?”
- “Is there a view for daily revenue in the analytics dataset?”
Every column’s name and type
A table or view’s columns — name, type and mode, with nested fields one level deep — plus its row and byte counts.
- “What columns does the orders table have?”
- “How many rows are in the sessions table, and how big is it?”
Also available: Projects available through your connection
1Create an API key
One key and one endpoint for every API on toolfare.
export TOOLFARE_API_KEY='tf_live_…'
2Every tool and its input
Send the tool name and its arguments as JSON; the reply is JSON too.
bigquery_list_datasetsEvery dataset the connection can see, with its region — the starting point for knowing what data exists at all.
bigquery_list_tablesTables, views, materialised views and external tables in one dataset, each with its type — views are often the curated entry point, and they are listed alongside tables.
bigquery_get_table_schemaA table or view’s columns — name, type and mode, with nested fields one level deep — plus its row and byte counts.
bigquery_list_projectsProjects available through your connection
Use it when
- You are about to write SQL and want the real column names rather than a guess.
- You want to know which dataset or view holds the numbers you are after.
- You are checking how large a table is before querying it.
Look elsewhere when
- You want the rows themselves. Running SQL needs a person to review the query first, and that step is not published through the MCP connection.
- You want marketing numbers that already have a source. Your ad and analytics accounts answer those directly. Use this instead
BigQuery directly, or through toolfare
Directly
- The BigQuery console
- A service account and key of your own
- Your own code to walk datasets and schemas
Through toolfare
- Ask in the chat you already use, in plain English
- One toolfare login; no service-account key handed to an agent
- Schema context beside the rest of your marketing data
You get the warehouse’s map, not its contents. No SQL runs and no rows come back through the MCP connection — for that you still need the BigQuery console or a reviewed query in toolfare.
Limits
- The connection lists datasets, tables and schemas; it does not run queries or return table rows.
- Listing a table does not prove it is readable — row-level permission is checked only when a query runs.
- Nested columns are expanded one level deep.
