Home/cubic
cubic AI code review logo

cubic

cubic reviews pull requests and entire codebases, learns team conventions, explains changes, suggests fixes, and runs scheduled agents to find bugs and vulnerabilities.

Visit website
cubic AI pull request review and codebase scanning platform

Product overview

What is cubic?

cubic is an AI code-review platform for pull requests and large codebases. It installs into GitHub and supported development environments, reviews changes in seconds, writes pull-request summaries, flags bugs and technical debt, and can apply straightforward fixes. Developers can also ask questions about a pull request or codebase.

The reviewer adapts to a team's standards through rules written in plain English, previous senior-engineer review comments, and context from tools such as issue trackers and documentation systems. Beyond pull-request review, codebase scans run many background agents to find bugs and security issues, notify owners, create tickets, and propose fixes on a schedule or before a release.

How to Use cubic

  1. Connect the relevant GitHub organization or repository.
  2. Select repositories and grant only the permissions required for review.
  3. Add team rules and connect approved project-management or documentation context.
  4. Open a pull request and inspect cubic's inline findings and summary.
  5. Accept, reject, or refine suggested fixes after human review and tests.
  6. Configure scheduled codebase scans for deeper bug and vulnerability discovery.

Core Features

  • Instant PR reviews: Adds inline findings to pull requests in GitHub and the IDE.
  • PR summaries: Explains the intent, impact, and logical groups of a change.
  • One-click fixes: Commits simple fixes and hands harder tasks to background coding agents.
  • Team learning: Uses explicit rules, prior review comments, and connected work context.
  • Codebase scans: Runs scheduled agents across repositories to find bugs and security issues.
  • Triage automation: Notifies owners, creates tickets, and closes them when fixes merge.
  • Privacy controls: The site states that code is not stored or used to train AI, and lists SOC 2 compliance.

Use Cases

  • Pull-request quality: Detect logic bugs, missed edge cases, and standards violations before merge.
  • Review acceleration: Summarize complex diffs and reduce repetitive reviewer work.
  • Release readiness: Scan a codebase before a major release.
  • Security discovery: Search repositories for potential vulnerabilities and assign remediation.
  • Engineering onboarding: Encode team conventions and connect project context for consistent reviews.

Pricing

The Starter plan is free and includes 20 PR reviews per month and up to five custom agents. Team costs $30 per developer per month when billed annually or $40 monthly. Pro costs $79 annually billed or $99 monthly per developer. Both paid plans offer a seven-day trial. Enterprise pricing is custom, and public open-source repositories can use cubic for free.

Frequently Asked Questions

Which languages does cubic support?

The official site describes it as language-agnostic and lists popular languages including JavaScript, TypeScript, Python, Go, Ruby, Java, and C#.

Does repository size determine usage?

No. Reviewed-line usage is based on pull-request changes cubic actually reads.

Should teams merge automated fixes without review?

No. Generated patches should pass the same human review, tests, and security checks as other code.

Back to product directory

Related products

OpenAI Codex is an AI coding product for delegating scoped repository tasks and reviewing the resulting changes.

Claude Code is Anthropic’s AI coding product for iterative repository work through supported development surfaces.

Gemini CLI is Google’s open-source AI agent for terminal workflows and local development tasks.

Newsletter

Keep up with useful AI products

Get a concise selection of new products, practical use cases, and builder updates.