CodeRabbit puts GPT-6.1 Sol in its security agent

What was announced?

OpenAI published a customer story on Monday (OpenAI): CodeRabbit uses GPT-6.1 Sol as the primary model in its security agent. The goal is to take review beyond individual changes and show risks across the whole codebase.

How does it work?

A pull request can affect shared libraries, dependencies and safeguards elsewhere in an application. Sol connects behavior across files and components to surface potential security issues. CodeRabbit says its review is independent of the agent that wrote the code and complements established scanners. The system also uses other frontier models and security tooling.

It works at three levels. During pull request review it analyzes changes as developers work. Deep Scan investigates whole repositories, including dependencies and how components interact. CodeRabbit Secure extends this to linked repositories, such as a shared authentication library, to show how a weakness could affect the wider system.

What does Secure add?

For user-selected findings, it can attempt exploit verification in an isolated sandbox without testing production systems. For supported issues, it can draft a pull request with a proposed fix for a person to review and approve. CodeRabbit says Sol has flagged security issues across thousands of repositories and organizations; the post gives no exact numbers or benchmarks.

What to keep in mind

This is a customer story, not an independent evaluation. Why it matters: if AI agents write part of your code, it makes sense to hand security review to a separate, independent system that also looks at effects beyond the PR. People still decide what ships.