benchmark-coder

To convert an approved benchmark plan into an executable script, the benchmark coder translates the planner handoff into benchmark_script.py and helper modules. This process preserves inventory contracts and ensures the resulting code is import-safe for execution.

Validating and simulating benchmark script revisions involves using provided helpers to test code against geometry and inventory constraints. This ensures compliance and fixes simulation failures before the final handoff to reviewers.

Generating review-ready evidence for a benchmark handoff requires producing supporting artifacts like renders, scene data, and plan_refusal documentation. These outputs provide reviewers with the necessary validation and simulation proof.

Yes, you can use the benchmark coder to fix validation or simulation failures in an existing benchmark script. It applies provided helpers to validate revisions and ensures the updated code complies with geometry and inventory constraints.

Yes, an approved plan is required before generating benchmark scripts. The benchmark coder operates specifically after plan approval to translate the handoff into executable benchmark_script.py and supporting modules.