query_execute
Read-only
Execute a Jaquel query directly on the connected ODS server.
Instructions
Execute a Jaquel query directly on connected ODS server.
Input Schema
TableJSON Schema
| Name | Required | Description | Default |
|---|---|---|---|
| query | Yes | Jaquel query dict to execute. Top-level key is entity name, value is filter/attribute object. Example: {"AoTest": {"name": {"$like": "*"}}, "$attributes": {"id": 1, "name": 1}, "$options": {"$rowlimit": 100}} | |
| max_rows | No | Maximum number of rows to return (default: 100). Also capped adaptively by max_cells to protect LLM context size. Use a small value like 10-20 for wide results (many columns). | |
| max_cells | No | Adaptive cell budget: effective_rows = min(max_rows, max_cells // col_count). Default 10 000 ≈ 6 000 LLM tokens for double data. Increase only if you need more data and understand the context cost. | |
| result_format | No | Result serialisation format: "split" (default) encodes column names once — {"columns": [...], "index": [...], "data": [...]}; "records" repeats all keys per row — [{"col": val, ...}, ...]. "split" is more token-efficient for wide results. | split |
Output Schema
TableJSON Schema
| Name | Required | Description | Default |
|---|---|---|---|
No arguments | |||