FAQ

What is agentic code review?

Other review tools send your change to a single model for one pass. Vetoo instead runs autonomous agents that read the request, gather the context around it, and act on what they find. Vetoo then reconciles their findings into one verdict. See The agent panel.

Which models does Vetoo use?

Every review runs a panel of agents, and the agents run on different models, so several models read each change. On Enterprise, you can run reviews on your own model key. See Bring your own key.

What happens to my code?

Vetoo encrypts your code in transit and processes it in isolation for the length of the review. Vetoo deletes the code once it posts the verdict. Vetoo never stores your source and never trains on it. See Security and data handling.

Which platforms does Vetoo support?

Two: GitHub and GitLab. Vetoo reviews pull requests on GitHub and merge requests on GitLab, and posts inline comments on both. GitHub connects through the GitHub App. GitLab connects over OAuth, or over a personal access token with the api scope. See GitHub and GitLab.

Which languages does Vetoo support?

All of them. Vetoo is language agnostic and reviews any language, from the most widely used to the most niche.

How does pricing work?

Every paid plan bills per seat. Each seat's credit allowance adds to one shared balance, and reviews spend from that balance. When you need more credits, buy them, or move to a higher tier. See Plans and credits.

Can I control what a review costs?

Yes. Review effort has five levels, from LOW at about 100 credits a review to ULTRA at about 1,500 credits. Set a default for the workspace. Then override it per repository. See Review effort.

Can I tune what Vetoo flags?

Yes, in two ways. Commit a VETOO.md file to set explicit rules. Vote findings up or down, and reply with comments, to calibrate a repository over time. See VETOO.md and Team memory.