Tooling

Benchmarking Model Prompt Tests

Introducing the first benchmark runner for model-prompt-tests: a small harness for running prompt suites across model providers, scoring outputs against rubrics, and publishing comparable reports.