# Provider API Harness

Owner: Benchmark engineer

## Input

Suite task packet plus immutable prompt packet hash.

## Output

One JSONL row per model run with model id, provider id, latency, token/cost fields, response artifact URI, and reviewer queue id.

## Required Evidence

- provider response id
- model version
- raw prompt packet
- raw answer artifact
- cost ledger

## Rejection Rule

A provider run is rejected when the model id, response id, cost, or prompt packet cannot be reproduced.

## Handoff

Normalize accepted rows into either the first-party benchmark run intake form or `public/reports/benchmark-intake/run-template.csv`, then run `pnpm benchmarks:generate` after reviewer signoff.