Skip to main content
critiqor runs prints a summary table of every finalized Critiqor evaluation run found in the local runs directory. For each run it shows the trust score, total tool call count, hallucination risk band, and the primary failure type identified by the diagnosis engine. Use it to get a quick overview of agent reliability across sessions, or to find a run_id before opening a specific dashboard.

​
Syntax

critiqor runs [OPTIONS]

​
Options

OptionDefaultDescription
--runs-dir TEXTrunsDirectory containing run artifacts

​
Example output

Available Runs

run_001 | Trust: 82 | 14 Tool Calls | Hallucination Risk: Low    | No Major Issue
run_002 | Trust: 61 | 31 Tool Calls | Hallucination Risk: Review | Infinite Tool Loop
run_003 | Trust: 47 |  8 Tool Calls | Hallucination Risk: High   | Memory Degradation
Runs are listed most-recent first. If no finalized runs are found, the command prompts you to run critiqor finalize.

​
Understanding the columns

ColumnDescription
Trust score0–100 weighted reliability score across 6 dimensions. Higher is better.
Tool CallsTotal tool calls captured and verified in the run
Hallucination RiskLow (trust ≥ 75) · Review (trust 60–74) · High (trust < 60)
Primary diagnosisThe most impactful detected failure type, or No Major Issue if none was found

​
Workflow

critiqor runs is the fastest way to find a run ID for further inspection: 
critiqor runs                      # see all completed runs
critiqor dashboard run_002         # open the full diagnosis for a specific run