Installation
Install Critiqor from PyPI and verify your setup in under a minute.
Quick Start
Observe your first agent run and open the dashboard end-to-end.
CLI Reference
Every command, flag, and option — searchable and ready to copy.
Dashboard Guide
Understand every section of the Critiqor reliability dashboard.
How Critiqor works
Critiqor fits into your existing workflow without changing how you use OpenClaw. Start an observation session, use your agent as normal, then finalize to get a reliability report.1
Start observing
Run
critiqor monitor openclaw before you use your agent. Critiqor starts capturing runtime evidence in the background.2
Use your agent normally
Work with OpenClaw as you usually would. Critiqor stays out of the way while recording what your agent actually does.
3
Finalize and review
Run
critiqor finalize when you’re done. Critiqor generates a reliability report and opens the dashboard automatically.Explore the docs
Browse the full Critiqor documentation to learn how runtime observation works, how to read your reliability reports, and how to get the most from every evaluation.Runtime Observation
Learn what Critiqor captures during an agent session.
Trust Score
Understand how to read and act on trust signals.
Observation Workflow
Step-by-step guide to starting, ending, and reviewing runs.
Best Practices
Get the most useful evaluations from every session.
Troubleshooting
Fix common installation and runtime issues fast.
FAQ
Quick answers to the most common questions.