Frequently Asked Questions＃

• Human-written code: Direct editor integration. You trigger reviews yourself by right-clicking code, using keyboard shortcuts, or clicking buttons in the extension sidebar. No AI assistant needed—review any code directly.
• AI-generated code: For AI-assisted workflows. Reviews trigger automatically when AI generates code, or when you ask your AI assistant to review existing code. Works with Claude Code, Codex CLI, Cursor, VS Code, and other AI assistants.

Yes. Human-written code reviews are built for this. Right-click any code or press Ctrl+Shift+K to trigger a review. No AI assistant needed. See Human-written code quickstart for setup.

Run an on-demand review before pushing changes. Use Review uncommitted changes for staged files, or Review all branch changes to review all changes made since the branch was created.

No. AI-generated code reviews run when AI generates code. Human-written code reviews run when you trigger them. They work independently.

• IDE extensions: Cursor, VS Code, Windsurf, Antigravity
• CLI tools: Claude Code, Codex CLI
• Standalone CLI: kluster-cli for terminal-only workflows

See Installation for setup instructions.

kluster.ai is language agnostic and can review code in any programming language, including Python, TypeScript, JavaScript, Java, Go, Rust, C++, C#, Ruby, PHP, and more.

Yes. For system-wide analysis, use Repo Reviews to scan your entire repository and find bugs that emerge from cross-module interactions. Repo reviews are designed specifically for analyzing your complete codebase as a system.

For reviewing specific changes, use on-demand reviews to check individual files, code blocks, or uncommitted changes.

Yes. When using an AI coding assistant, Code Reviews works with any model available in supported IDEs, including Claude, GPT, Gemini, and others.

Currently, exclusion rules are configured at the project level through the kluster.ai platform. See Custom Rules for setting up project-specific configurations.

Team management is available on Team and Enterprise plans. Log in to the kluster.ai platform and navigate to your team settings to invite members.

Activation codes provide promotional credits for kluster.ai plans. See our detailed guide: How to use activation codes.

kluster.ai sources vulnerability data from public CVE databases that are continuously updated. Detection coverage improves as new CVEs are published and ingested, rather than following a fixed update schedule.

Each time a code review is done, a feedback option is available from the extension or in the platform. Your feedback helps improve detection accuracy and reduce false positives.

Yes. When you connect your GitHub repositories, kluster.ai learns rules from your repo and applies project-specific configurations. See Custom Rules for more details.

We also continuously improve our engine to perform deeper reviews, optimizing for common issues our users encounter. If you have suggestions for improvement, contact us at support@kluster.ai.

• Instant: Completes in about five seconds. Detects most issues and is suited for fast review and iteration cycles.
• Deep: Takes up to a few minutes. Conducts deeper analysis to uncover even subtle edge cases, ideal for critical code and final reviews.

Use Instant for fast iteration during development. Use Deep for final reviews, production code, or when thoroughness matters more than speed. You can configure the default for each workflow in Options.

For AI-generated code, reviews trigger automatically when your AI coding assistant generates or modifies code. This happens without any action from you. The code diff is sent to kluster.ai, and the results appear in your chat or terminal.

AI-generated code reviews see your original prompt to the AI, so they can verify the AI did what you asked—not just that the code runs. If you asked for Firebase auth but the AI used localStorage, the review catches it. Human-written code reviews cannot check intent because they only see the code, not your request. See this example for a real case.

No. kluster.ai identifies issues and suggests fixes, but doesn't modify your code directly. For AI-generated code, the AI assistant decides whether to apply them—in most cases, it will automatically incorporate the fixes. For human-written code, you apply fixes yourself using the Fix with AI button or manually. Either way, you remain in control.

Yes. You can disable automatic reviews from Options in the kluster.ai platform. Alternatively, you can disable the kluster.ai extension in your IDE, or for CLI tools, disable the MCP server.

Yes. AI-generated code reviews analyze diffs across multiple files in a single review, understanding the context of changes that span your codebase.

The Dependency Validator automatically checks packages before installation. When your AI assistant suggests adding a dependency, kluster.ai validates it for known vulnerabilities before the install command runs.

The changes are made by your AI assistant, not kluster.ai itself. You can revert changes in the file using your IDE undo functionality, Git commands, or simply ask the AI assistant to revert the changes.

Select the code, press Ctrl+Shift+K (or right-click → Review with kluster.ai). Results appear in seconds with issues and one-click fixes.

Yes. Check out the PR locally and either:

1. Use Review uncommitted changes to scan all modifications.
2. Select specific code and review it.

You can trigger an on-demand review in three ways:

1. Right-click selected code → Review with kluster.ai.
2. Use the extension sidebar. Open the On-Demand Review section, select a mode, and click Instant Review or Deep Review.
3. Select code and use the hint button. This option is not available in Cursor yet.

Yes. On-demand reviews work on any code, including your own code, AI-generated code, or code from colleagues. This is useful for auditing existing codebases or reviewing pull requests.

Yes. Select the code you want to review, then right-click and choose Review with kluster.ai. You can review anything from a single function to an entire file.

Yes. In the kluster.ai extension sidebar, open the On-Demand Review section, select Review uncommitted changes from the Mode dropdown, then click Instant Review or Deep Review. This reviews all staged and unstaged changes across your repository.

• On-demand reviews: You click a button or use a shortcut to trigger a review.
• Background auto reviews (Beta, Enterprise plan): Automatically review your code for issues and suggestions as you work, without requiring you to trigger anything. Enable it from the Enabled Tools section in Options.

Yes. The kluster-cli tool provides full code review functionality directly from the terminal—no IDE or AI assistant needed. Install it, authenticate with your API key, and review staged changes, diffs, or individual files. See CLI Quickstart.

Install git hooks with kluster hooks install pre-commit|pre-push|all. Choose pre-commit to review before every commit, pre-push to review before pushing, or all for both. See Git Hooks.

Yes. Use --output json for machine-readable output and check exit codes to fail builds on issues. Exit code 0 means no issues, 3 means high severity, and 4 means critical. See Reference for the full table.

Yes. Use kluster review file <path> to review any file, even outside a git repository. See Review Commands.

Repo reviews analyze your entire codebase as a system instead of reviewing individual changes. They find bugs that emerge from interactions across modules—issues that don't show up in PR-level reviews because they're only visible when you examine how multiple parts of your code work together.

Other review modes (on-demand and automatic) check specific changes you make. Repo reviews scan everything to catch cross-module bugs, state inconsistencies, bypassed validation checks, and system-wide problems.

Depending on repository size, repo reviews take several minutes. You'll receive an email notification once the analysis completes. You can close the page and come back later—the review continues running in the background.

Yes. Repo reviews work with both public and private repositories on GitHub, GitLab, and Bitbucket. The connection is secure and respects your repository permissions.

Repo reviews are available only through the web dashboard at platform.kluster.ai/repo-reviews. You need to connect your GitHub, GitLab, or Bitbucket account first. Repo reviews are not available in IDE extensions or CLI tools.