Schema Load
A reproducible CLI for testing whether OpenAI-compatible model services still return correct JSON under load. The runner varies concurrency, prompt and output length, temperature, seed, and structured-output mode.
- Role
- Evaluation tooling and systems engineering
- Focus
- Structured-output reliability under load


