ct run monitor

Run monitors on trajectories (action-level and/or trajectory-level)

Usage

ct run monitor [OPTIONS] TRAJ_LIKE

Arguments

ArgumentTypeRequired
TRAJ_LIKEtextYes

Options

OptionDescription
-a, --action-monitor TEXTAction monitor selection: a name, a .md asset path, a module:function import path, or a JSON list of entries (each a selector or a {"monitor": ..., <overrides>} mapping). ct protocols monitors action to see available. [default: mean-action]
-t, --trajectory-monitor TEXTTrajectory monitor selection: a name, a .md asset path, a module:function import path, or a JSON list of entries (each a selector or a {"monitor": ..., <overrides>} mapping). ct protocols monitors trajectory to see available. [default: mean-quick]
--trusted-model TEXTModel bound as the eval-level trusted role the monitors run on. [required]
--only-actionOnly run the action monitor
--only-trajectoryOnly run the trajectory monitor
-o, --save-dir PATHDirectory to save monitored trajectories (default: ./data/monitoring/{monitor}/{traj_name}_{timestamp})
--upload / --no-uploadUpload results to the Control Tower database (enabled by default). [default: upload]
--run-name TEXTName for the run (default: monitor-{monitor_name})
--tag TEXTTags for the run (can be specified multiple times)
-L, --limit INTEGER RANGELimit the number of trajectories to monitor [x>=1]
-C, --max-concurrent INTEGERMax trajectories to monitor concurrently (default: unlimited)
--cot-level [none|reasoning|extended]Chain-of-thought visibility for monitors: none, reasoning (models output text), or extended (reasoning and CoT). [default: none]
--helpShow this message and exit.