call_api
ChatGPTQuery the Agentio Brand API — Agentio's own first-party API for a brand's reporting data. Calls one of the read-only operations advertised by list_operations and returns the JSON response. Every call is automatically scoped to the brand's own data. Discover operations with
list_operations, and read the operation's doc page (api/<operation_id>.md via read_product_doc) before the first call so parameters and response fields are used correctly. An empty result answers only exactly what you queried — never read it as proof the data doesn't exist elsewhere. Reporting rows show activity in the queried window, not current campaign status. Zero, null, and no rows are distinct; none alone diagnoses a tracking or ingestion fault. Interpretation: read this operation's cost definition, date basis, and denominator before choosing labels or aggregating. When a field represents full contracted cost, label it contracted cost in headlines and tables, not period spend or invoices. Check whether costs repeat across rows or windows before summing; different reporting families can use different cost bases. Use supplied rollups for totals and ratios. Read the operation's documented metric and missing-value meanings before interpreting a result or selecting a proxy. This tool supports rankings and explanations of recorded changes. Renewal, budget, pricing, and test-plan decisions belong with Ask Agentio in the Agentio dashboard after presenting the relevant numbers. Arguments: operation_id; path_params — object filling the {placeholders} in the operation's path (values must be UUIDs from earlier responses); query_params — object of documented query parameters (optional).
