```json { "table_name": "llm_calls", "table_type_tags": [ {"tag": "dimension", "confidence": 0.90} ], "primary_type": "dimension", "insufficient_evidence": false, "entity": { "columns": ["id"], "unique_per_row": true }, "time_axis": { "columns": ["created_at"], "kind": "event_time" }, "reasoning": "The entity id is unique per row (Uniqueness 1.000, RPUV 1.00), making each row a distinct LLM API call record. Although created_at is a timestamp (Uniqueness 1.000), it is an attribute of when each call occurred, not an axis ordering state transitions of a repeating entity. The table catalogs individual API calls with their properties (purpose, model, tokens, cost, latency, errors) — each row IS a complete call record, not an action that happened TO something or a snapshot OF something. This is a catalog of occurrences with unique identifiers: dimension." } ```