AgentnessBench / tests /agents /test_vanilla.py

Commit History

refactor: restructure proteus into game/web subpackages
426093b

irregular6612 Claude Opus 4.8 (1M context) commited on

feat(cp4.5): preserve probe reasoning + raw_text + token accounting via ProbeResult
f95852b

irregular6612 commited on

feat(cp4.5): persist act-call token accounting + raw_text in TurnTrace
1c18fa9

irregular6612 commited on

fix(cp2): VanillaAgent parses action from think-stripped answer (ignore in-think decoys)
f8966db

irregular6612 commited on

feat(cp2): VanillaAgent (act + probe, reasoning capture)
5f1a029

irregular6612 commited on