Vibeleaderboard
Index / tool
Visit www.critic.run
Category
Developer Tools
Rank
No. 4093Tools index

Previous survey · No. 3866 ·

Type
TOOL
Use case
Software Testing & Quality
Interfaces
MCP · Agent Skill / Plugin
Date

About

Critic is a code review platform from Feyn that lets reviewers talk directly with the AI agent that wrote a change instead of just reading the diff. A Codex or Claude Code plugin has the agent write a narrative behind each change and annotate key blocks with evidence like screenshots, and every feature is exposed over MCP so a reviewer's question forks the original authoring session and forwards it there, keeping the main thread clean.

At a glance

Critic puts a change narrative, diffs and reviewer questions together for agent-written code. It is worth a pilot if reviewers regularly need to ask why a change was made and the author’s coding session has context missing from the pull request.

Good fit

GitHub teams reviewing agent-generated changes that need explanation alongside the diff.

Consider something else when

Skip it if review patches and local change snapshots must stay off a hosted service.

Start here

Use a disposable repository and one small, known change. Compare the narrative with the actual diff, then ask a question whose answer you can verify in the code.

Cost and requirements

The product offers GitHub sign-in and uses a GitHub App for selected repositories. The local bridge supplies change context. Check plan terms before rolling it out to a team.

Limits to know

The policy says Critic receives patches, change snapshots and review discussion while raw transcripts stay local. Asking a connected agent a question can send relevant review and repository context to that agent’s provider.

Still unclear

  • What plan limits and prices apply to a team rollout?

Sources checked 2026-11-05. Source-based assessment.

Why it made the leaderboard

Instead of reverse-engineering intent from a diff, Critic lets reviewers chat directly with the agent that wrote the change, forking its original session so you get the actual reasoning behind a decision rather than a guess.

Intel on Critic

More in Intel

Tags

code-reviewai-agentsmcpclaude-codecodexpull-requests

Comments (0)

No comments yet

Editorially curated, with community endorsements as a secondary signal. Corrections welcome.