Skip to content
Critique

About Critique

Built for code that has to hold.

AI agents write code. Critique verifies it in a sandbox before you merge. Critique is an Automated code verification platform for GitHub pull requests — not another comment bot on your diff.

We gate untrusted changes, run evidence-backed review, enforce merge policy, repair with proof, and expose the same control plane over the Dashboard, GitHub, and Platform API — one Change Passport per pull request.

Change Passport

One governed record per pull request

Merge boundary

Gate, evidence, policy, repair

Platform API

Same control plane everywhere

Create accountOpen WorkspaceRead docs
Critique product surfaces: review board, architecture map, and repository chat
Review, map, and chat share one repo context — not three disconnected tabs.

Change control

Not just a code review CI tool.

Critique treats the Change Passport as the product object: provenance, risk, evidence, policy decisions, remedy proof, and memory on one merge-boundary record. Review runs are commit-level evidence generators inside that passport.

Linked context

Chat, review, and remedy share repository state instead of re-explaining the PR.

Checkpoint first

Policy gates run before the review swarm spends on untrusted shape.

Visible artifacts

Findings, scope, and handoff contracts stay on the trail.

Bounded execution

Automation stops at agreed scope — not an open-ended rewrite.

Merge-boundary phases

Phase 1

Gate

When
Before the review queue
Check
Critique / Checkpoint
What it does
Fast, cheap firewall rules

Phase 2

Review

When
During the run
Check
Critique / Review
What it does
Multi-model specialists + lead verdict

Phase 3

Merge

When
After evidence
Check
Critique / Merge Policy
What it does
Enforceable merge rules with proof
Change control docsPR control guideControl BoardHow Critique works

The loop

Chat, review, then repair.

We started with review because finding the truth has to come before asking automation to act. Checkpoint can gate the PR before the review swarm spends.

Chat

Ask repository questions without starting from zero.

Review

Run a grounded PR critique with specialist lanes.

Remedy

Turn eligible findings into validated branch updates.

Checkpoint modes

  • dry-run

    Observe policy impact without blocking contributors.

  • warn

    Surface risk while still allowing the review to continue.

  • block

    Stop untrusted or malformed PRs before full automation starts.

See Checkpoint
Read docsVersion history (v3.1)

Platform

One merge boundary. Many surfaces.

Critique is not a single feature — it is a control plane for how AI-authored changes enter your codebase. Each surface below ships in beta and links to operator docs plus the ship log entry where it landed.

Full index: documentation · version history · Skill Marketplace

Principles

How we decide what ships.

  1. Evidence over vibes

    trace

    A finding should carry enough context for an engineer to verify it fast.

  2. Trust is earned

    gate

    Automation must show its work, obey scope, and accept when it should stop.

  3. Code before theatre

    quiet

    The interface stays quiet because the work is already complicated.

  4. Taste under pressure

    craft

    Developer tools should feel precise even when the system is busy.

  5. Local control

    policy

    Teams choose policies, models, and where execution runs.

Autonomy with boundaries. That is how trust is earned.

Shipping now

Live in beta, shipping weekly.

The list below is the short operator view. Each row in the Platform section above links to docs and the release that introduced it.

  • Change Passports — provenance, risk, evidence, policy, and remedy proof per PR
  • Control Board — gate, policy, memory, delivery, and learnings in one operator surface
  • GitHub App review runs with evidence contracts and specialist lanes
  • Agent Workspace — chat, cloud lanes, builder, and repair on one shell
  • Checkpoint (Agent Firewall) on the merge boundary
  • Remedy with proof bundles and bounded sandbox execution
  • Critique Intake — agentic bug intake from user report to handoff
  • Platform API v1 — queue reviews, passports, webhooks, and MCP
  • Merge Gate API, Inference API, and Coding Agent API
  • Agent Skill Marketplace with performance leaderboard
  • BYOK (OpenRouter / CrofAI) and BYOA (Cursor, Claude, Codex)
  • Insights — velocity, cost, retrospectives, and compliance exports

Read docs · Version history · Skill Marketplace

Start here

Better code. Shipped with confidence.

Connect GitHub, run your next PR through review, and use Remedy when the fix has enough evidence to act.

Create accountSee Remedy

Contact

Product and account help: support@critique.sh. Founder: Repath Khan (ray@critique.sh).