Local pull-request diff review
Reviews pull-request diffs against the complete local checkout, with changed-file tracking, multi-file navigation, and observations displayed beside the relevant code.
CodeCrab
CodeCrab is a native desktop app for reviewing GitHub pull requests against a local repository checkout. It helps engineers inspect inline AI findings, apply repository-specific review skills, and keep source code and review execution on their own machine.
Local-first pull-request review
CodeCrab is a native desktop application for reviewing GitHub pull requests against a local repository checkout. It loads the diff locally, shows changed files and inline AI observations, and uses the repository's code, rules, and selected review skills to investigate potential correctness, security, and regression issues.
The product is designed for engineers who want local execution and control over the review process. It runs on the user's machine, does not upload source code to a vendor cloud, uses existing local AI tooling such as Claude Code, and remains read-only today: users decide which findings to act on, comment on, or fix.
Features
Repository-aware profiles
Builds a repository-specific review profile from project rules and patterns, rather than applying only a generic checklist.
Configurable review skills
Supports local skills and specialized review skills for areas including security, test completeness, and correctness; users can choose which skills to run.
Interactive findings workflow
Shows a live execution console and lets engineers investigate findings, ask follow-up questions, and inspect the reasoning in context.
Local AI tool orchestration
Uses existing local tools and subscriptions, including Claude Code and the site's mentioned Cursor workflow, while allowing users to choose their AI models and control review spending.
Local test verification and control
Runs native tests such as cargo test, pytest, or npm test before changes are applied, while keeping the workflow read-only by default.
Use cases
Team pull-request review
Review a teammate's pull request with the full checkout in context, navigate changed files, investigate inline findings, and decide which issues warrant action before merge.
Pre-push quality checks
Run a review on local changes before they become a pull request, using repository rules and a pre-push skill to catch issues earlier in the development cycle.
Clarifying feedback on your own PR
Investigate vague or unfamiliar review comments by tracing callers, tests, neighboring modules, and the surrounding GitHub conversation, then produce an editable fix prompt for a local AI tool.
Risk-focused review
Apply a deeper, targeted review skill to sensitive changes such as authentication work, while using lighter criteria for routine changes.
Pros and Cons
Pros
- Runs reviews on the local machine without uploading source code to a vendor cloud.
- Uses the full local checkout, repository-specific profiles, and selectable custom skills rather than relying only on the pull-request diff.
- Provides inline observations, a live execution console, follow-up investigation, and multi-file diff navigation.
- Works with local development context and can run native test suites such as cargo test, pytest, and npm test.
- Keeps engineers in control by avoiding commits, pushes, and public GitHub comments by default.
Cons
- The available evidence describes a free public beta and does not establish paid-plan details, usage limits, or long-term availability.
- The product is read-only today; automatic fixes, publishing review comments, and other write actions are described as future or planned workflows rather than current features.
- Setup requirements and the full list of supported platforms and integrations are not specified in the available product content.
FAQ
How does CodeCrab review a pull request?
CodeCrab runs as a native desktop application and performs reviews against the local checkout. The source describes a GitHub workflow using your existing local gh login, but it does not provide complete setup instructions.
Does CodeCrab upload source code to a server?
The product is described as running 100% on the user's machine, with source code not uploaded to a vendor cloud. It uses local tools and subscriptions such as Claude Code, and the site also mentions Cursor.
Does CodeCrab change code or post review comments automatically?
CodeCrab is read-only today: it does not commit, push, or publish a public GitHub review comment without the user's action. The product describes automatic fixing as a future opt-in workflow rather than a current capability.
Can reviews use repository-specific rules and skills?
It can use a repository profile, local review skills, and specialized skills for areas such as security, test completeness, and correctness. Users can choose which skills to include for a review.
How much does CodeCrab cost?
The site describes CodeCrab as a free public beta. The pricing page currently returns a page-not-found response, so no paid plans, prices, or usage limits are established by the available evidence.
Quick Facts
- Category
- Developer Tool
- Primary workflow
- Local GitHub pull-request review
- App type
- Native desktop application
- Execution model
- 100% on the user's machine
- AI tooling
- Claude Code and local skills; Cursor is mentioned on the product site
- Availability
- Free public beta
Alternatives à CodeCrab
ByteAsk
ByteAsk is a terminal-first AI coding agent for C and C++ that edits repositories and verifies changes with the real compiler, debugger, sanitizers, and tests before showing a diff. It offers a free tier plus paid plans, with editor connectors and zero-retention handling described in the source.
GitBot
GitBot is a local browser-based workspace for creating reusable coding bots powered by Claude Code, Codex, or OpenCode. It helps developers run those agents across repository folders with persistent threads and configurable permissions.
Manta AI
Manta AI is an autonomous web app testing tool for teams that want to map application behavior, catch regressions, and generate tests without writing scripts or maintaining selectors. It works from a URL and supports plain-English test flows, run results with screenshots, and scheduled or deployment-triggered checks.
Ghost
Ghost est un assistant IA en terminal pour discuter, générer du code et lancer des tâches en ligne de commande. Modèles gratuits, Linux, macOS, Windows, open source.
Vi3ecode
Vi3ecode is a maintained development environment for working with AI coding agents. It keeps project context, terminal, Git, memory, flows, review, and communication connected around the agent you already use.
Lucid Train
Lucid Train is a local-first AI coding harness that turns repository code into architecture diagrams and uses those diagrams as specifications for coding agents. It is available as a desktop app and a lightweight Rust CLI, and it can run with local models or your own API keys.