Agent Arena

Reference

Precise lookup: CLI, config schema, scorers, APIs, glossary.

Precise lookup material. Completeness over prose.

Page Covers
cli.md Every command and flag
config-schema.md Every config.yaml and tests.yaml field
scorers.md The ten eval types, their options and failure modes
http-api.md The local JSON API
python-api.md Using the arena as a library
glossary.md Every term used in a specific way