We are thrilled to announce that the Cortex Model Context Protocol (MCP) is now available in the Devin marketplace. This integration connects the world’s first AI software engineer with the real-time context of your entire engineering ecosystem, as managed and measured by Cortex.
The rise of AI software engineers like Devin fundamentally changes how organizations tackle their biggest technical challenges. Teams can now delegate large-scale code refactors, migrations, and backlog tasks that previously took years of manual effort. But even the most powerful AI needs context. It needs to understand you; your services, your standards, and your definition of "good."
Supercharging Devin with your engineering standards
Once defined, Devin excels at executing complex engineering tasks at scale. Cortex excels at defining and measuring engineering standards.. Together, they form a context rich extension of your engineering team:
Cortex identifies the "what": Your Scorecards continuously evaluate your services against your standards for production readiness, security, reliability, and more, creating a prioritized list for Devin of what needs to be fixed.
Devin provides the "how": As an AI software engineer, Devin can take that list from Cortex and autonomously perform the work—creating PRs, updating documentation, and refactoring code—to bring services into compliance.
Without Cortex, an AI agent can’t know which of your services are using a deprecated library or failing a critical security check. With the Cortex MCP, Devin has the secure, context-rich data it needs to work on your most important priorities, guided by the standards you've already set.
From Scorecards to automated PRs
Here’s a scenario that will likely sound familiar to you. A standard rule on a "Production Readiness" Scorecard is that every service must have a README.md file. This documentation is crucial for both human developers and AI tools.
Instead of your platform team manually chasing down service owners, you can now simply instruct Devin with the prompt below:
"Use the Cortex MCP to find all services failing the 'has a README' rule on the Production Readiness Scorecard, and open PRs to add a README file explaining what each service does."
Devin will connect to Cortex, retrieve the list of non-compliant services, and get to work. It systematically opens PRs, turning a passive Scorecard into an active, automated system for improving your engineering quality and eliminating toil.
Want to see this in action? Check out the demo in the video below:
This is just the beginning. You can use the same workflow to empower Devin to tackle more complex challenges, including:
Accelerating migrations: Use a Scorecard to track a library or cloud provider migration and have Devin refactor the code for all non-compliant services.
Remediating security vulnerabilities: When a new CVE is announced, create a Scorecard to identify affected services and task Devin with creating PRs to apply the patch.
Paying down tech debt: Use a Scorecard to identify services using a deprecated internal library and have Devin handle the upgrades.
Ready to connect the world's first AI software engineer to your engineering standards? Learn more and schedule a demo of Cortex today.
