# Critiqor

## Docs

- [Observation Engine: Session Lifecycle and Evidence Capture](https://critiqor.mintlify.app/architecture/observation-engine.md): Critiqor's observation engine manages run sessions from creation through finalization, capturing runtime events and orchestrating diagnosis generation.
- [OpenClaw Plugin: Hook-Based Runtime Instrumentation](https://critiqor.mintlify.app/architecture/openclaw-plugin.md): The Critiqor OpenClaw plugin hooks into tool lifecycle and extension API events to capture runtime evidence without modifying agent code or parsing logs.
- [Critiqor Architecture: CLI, Observation Engine, and Plugin](https://critiqor.mintlify.app/architecture/overview.md): Critiqor separates CLI routing from runtime supervision. A bundled plugin captures OpenClaw evidence; the observation engine generates diagnosis artifacts.
- [Dashboard Overview: Understanding Your Diagnosis Results](https://critiqor.mintlify.app/concepts/dashboard.md): The Critiqor local dashboard shows trust scores, failure diagnoses, cost analysis, causal graphs, and reliability benchmarks for every OpenClaw agent run.
- [Diagnosis Engine: From Evidence to Actionable Findings](https://critiqor.mintlify.app/concepts/diagnosis-engine.md): Critiqor's diagnosis engine scans session evidence with deterministic detectors to produce trust scores, failure causes, and deployment readiness verdicts.
- [Evidence Collection: How Critiqor Captures Runtime Data](https://critiqor.mintlify.app/concepts/evidence-collection.md): Critiqor captures tool calls, tool outputs, provider events, memory events, and execution metadata into a structured session file for every OpenClaw run.
- [Runtime Intelligence: Observe What Agents Actually Do](https://critiqor.mintlify.app/concepts/runtime-intelligence.md): Runtime intelligence evaluates AI agents from live execution traces, not final responses. Observed runtime evidence produces more reliable diagnoses.
- [Critiqor FAQ: Common Questions About Runtime Reliability](https://critiqor.mintlify.app/contributing/faq.md): Answers to the most common questions about Critiqor: how it works, data handling, run storage, dashboard usage, and framework support.
- [Critiqor Contributing Guide: Dev Setup and Workflow](https://critiqor.mintlify.app/contributing/setup.md): Clone the Critiqor repo, install in editable mode, run the test suite with pytest, build with python -m build, and publish a new release to PyPI.
- [Install Critiqor to Monitor OpenClaw Agent Sessions](https://critiqor.mintlify.app/getting-started/installation.md): Install Critiqor with pip on Python 3.10 to 3.13. Supports macOS, Linux, and Windows. Currently integrates with OpenClaw agents only.
- [Get Started with Critiqor: Monitor OpenClaw in Four Steps](https://critiqor.mintlify.app/getting-started/quickstart.md): Monitor your first OpenClaw session with Critiqor. Four commands: install, monitor, run, finalize — then your diagnosis dashboard opens automatically.
- [Local Dashboard: View and Understand Diagnosis Results](https://critiqor.mintlify.app/guides/dashboard.md): Critiqor's local dashboard reads diagnosis files from your runs directory. Learn how it launches, how to switch between runs, and what each section shows.
- [Evaluation: How Critiqor Scores and Diagnoses Agent Runs](https://critiqor.mintlify.app/guides/evaluation.md): Critiqor computes trust scores, failure diagnoses, causal graphs, and recommendations from runtime evidence using six deterministic dimension detectors.
- [Evidence Types: What Critiqor Collects During Agent Runs](https://critiqor.mintlify.app/guides/evidence.md): Detailed guide to the seven evidence categories Critiqor records during OpenClaw runs, with JSON examples and diagnostic relevance for each type.
- [Run Management: Store, Review, and Compare Agent Runs](https://critiqor.mintlify.app/guides/run-management.md): Critiqor stores every session as immutable run artifacts. List runs, switch between dashboards, and revisit any previous diagnosis using run IDs.
- [OpenClaw Integration: Full Agent Monitoring Setup and Usage](https://critiqor.mintlify.app/integrations/openclaw.md): Critiqor integrates with OpenClaw via a bundled plugin that captures runtime events automatically. No code changes to your agent are required.
- [Critiqor Integration Roadmap: Planned Framework Support](https://critiqor.mintlify.app/integrations/roadmap.md): Critiqor supports OpenClaw today with full plugin and dashboard integration. Claude Code, Hermes, and other frameworks are planned for future releases.
- [Critiqor: Runtime Intelligence Platform for AI Agents](https://critiqor.mintlify.app/introduction.md): Critiqor evaluates AI agents by observing their runtime behaviour, not their self-reports. Diagnose failures, build trust, and deploy with confidence.
- [Critiqor: Evidence-First Reliability Wrapper for AI Agents](https://critiqor.mintlify.app/reference/api/critiqor-class.md): Wraps any AI agent with evidence-first reliability evaluation, returning a 0-100 trust score, dimension critique, and deployment gate on each run.
- [Critiqor Data Types: Public API Type Reference Guide](https://critiqor.mintlify.app/reference/api/data-types.md): Reference for all Critiqor public data types: ToolCall, ToolOutput, FailureCause, EvaluationRecord, PolicyCheckResult, TrendAnalysis, and more.
- [EvidenceRecorder: Capture Tool Calls for Richer Scoring](https://critiqor.mintlify.app/reference/api/evidence-recorder.md): EvidenceRecorder is a context manager for manually capturing tool calls, tool outputs, and LLM calls to provide fully_instrumented evidence to Critiqor.
- [OpenClawRuntimeObserver: Python API for OpenClaw Evidence](https://critiqor.mintlify.app/reference/api/openclaw-observer.md): OpenClawRuntimeObserver and monitor_openclaw_process() record OpenClaw runtime events and build run payloads with trust scores and failure diagnosis.
- [critiqor dashboard — Open the Local Diagnosis Dashboard](https://critiqor.mintlify.app/reference/cli/dashboard.md): critiqor dashboard opens the local Critiqor dashboard for the latest or a specified run. Pass a run_id argument to view any previous run.
- [critiqor finalize — Generate Diagnosis and Open Dashboard](https://critiqor.mintlify.app/reference/cli/finalize.md): critiqor finalize stops an active monitoring session, runs the diagnosis engine, saves a diagnosis report, and launches the local dashboard.
- [critiqor monitor openclaw — Start Agent Observation](https://critiqor.mintlify.app/reference/cli/monitor.md): critiqor monitor openclaw launches OpenClaw with runtime observation. Options: --agent-id, --runs-dir, --difficulty-tier, --timeout, --visibility.
- [Critiqor CLI Commands: Full Reference and Workflow](https://critiqor.mintlify.app/reference/cli/overview.md): Complete reference for all Critiqor CLI commands. Use monitor, finalize, dashboard, and runs to observe OpenClaw agents and review diagnosis results.
- [critiqor runs — List All Completed Evaluation Runs](https://critiqor.mintlify.app/reference/cli/runs.md): critiqor runs lists all completed Critiqor evaluation runs with trust scores, tool call counts, hallucination risk, and primary diagnosis for each run.