list_tables
Lists available Dataroom tables and their schemas. Returns a compact summary by default. Setverbose to true to include full column details (name, data type, description).
| Parameter | Type | Required | Default | Description |
|---|---|---|---|---|
datasets | string[] | No | ["rp_data_external", "rp_data_intermediate"] | Filter by dataset name |
verbose | boolean | No | false | When true, includes column details (name, data type) |
rp_data_external, rp_data_intermediate
Sample prompt: “What tables are available for querying?“
query
Execute a read-only SQL query against the Dataroom. Use this for financial analytics — revenue breakdowns, cost reports, statement-level data, and period comparisons.| Parameter | Type | Required | Default | Description |
|---|---|---|---|---|
sql | string | Yes | - | SQL query with @project_id named parameter |
projectId | UUID | External auth only | Auto-injected | Project ID (required for external auth, auto-injected for internal) |
offset | integer | No | 0 | Number of rows to skip. Use with truncated results to paginate (e.g., offset=1000 for the next page) |
limit | integer | No | 1000 | Max rows to return in this response (hard-capped at 1,000). Lower it (e.g. 50–200) for wide schemas or SELECT * to keep the response small. |
columns | string[] | No | - | Optional allow-list of column names to keep in each returned row, applied after the query runs to shrink wide rows. Unknown names are ignored; omit to return all selected columns. |
truncated will be true in the response. Use offset to paginate through large result sets.
Sample prompt: “What are the top 10 recordings by revenue?”