Crit: Local-first, single-binary CLI for reviewing agent plans and diffs
Crit is a local-first review tool packaged as a single binary. It opens your file or code diff in a browser with a GitHub-inspired interface, lets you leave comments, then your agent acts on the feedback automatically. Repeat until satisfied — all before hitting GitHub.
Core loop: comment, agent updates, diff refreshes
Drag across a range of lines, leave comments, hit Finish Review. The agent picks up the feedback, edits the file, and Crit reloads with the updated diff. Previous comments stay visible; resolve only when you're satisfied. Works for both plans and code — same UI, beautifully rendered markdown with syntax highlighting for 190+ languages.
Install in one command
$ brew install crit $ go install github.com/tomasz-tomczyk/crit@latest $ nix profile install github:tomasz-tomczyk/crit # Windows: PS> iwr https://github.com/tomasz-tomczyk/crit/releases/latest/download/crit-windows-amd64.exe -OutFile crit.exe
Then run crit or crit plan.md. Pre-built binaries for ARM64, WSL also available.
Integrates with 11 agents
One-command setup for Claude Code, Cursor, Copilot, Codex, OpenCode, Gemini, Qwen, Hermes, Windsurf, Cline, Aider. Example:
$ claude plugin add tomasz-tomczyk/crit $ claude plugin install crit@crit $ crit install cursor $ crit install github-copilot $ crit install codex
See full setup docs for each agent.
Multi-round and concurrent reviews
Iterate over multiple rounds per file. Run several reviews in parallel without stepping on each other — useful when juggling multiple agents.
Optional sharing and self-hosting
Reviews stay local by default. One click generates a public link — no install, no login for viewers. You can self-host the share target (crit-web) for privacy. Without sharing, reviews auto-delete after 30 days of inactivity.
Small quality-of-life features
- Vim keybindings
- Git diffs (split & unified)
- Round-over-round diffs
- Mermaid diagrams
- Table of contents
- Dark & light themes
- Draft autosave
- Insert suggestions
Who it's for
Developers who frequently review AI-generated plans and code diffs and want a dedicated local UI to batch feedback and iterate without leaving their workflow.
📖 Read the full source: HN AI Agents
👀 See Also
Persistent AI Advisor with Cross-Platform Memory: Tracks Decision History for 3 Months
A Reddit user built a persistent AI advisor that remembers every product decision across Claude Code, Cursor, and a web interface, catching contradictions and improving over months.
Atlas Inference Engine Goes Open Source: Pure Rust + CUDA, 100+ tok/s on DGX Spark
Atlas is now open source — a Rust + CUDA inference engine that achieves 130 tok/s peak on Qwen3.5-35B (NVFP4) on a single DGX Spark, with no Python runtime and <2 minute cold start.
Claude Code UltraPlan Workflow Changes and Performance Observations
Claude Code UltraPlan introduces a cloud-based planning workflow with terminal launch, browser review interface, and execution options. Testing showed approximately 2x faster repeated runs than local planning, with mixed quality improvements.
repo-mem: Open-Source MCP Server Adds Persistent Team Memory to Claude Code
repo-mem is an open-source MCP server that adds persistent, shared memory to Claude Code sessions using SQLite and Git. It solves team isolation by storing observations in per-user databases that get committed to the repository.