Redgate Flyway Enterprise MCP for AI and Agentic Workflows

Huxley Kendell demonstrates how Redgate Flyway Enterprise enables safe, governed, and deterministic integration of AI and agentic workflows into database development processes. 0:00 Introduction to AI workflows 1:52 Local database development 4:23 Using MCP for database tasks 10:45 Fully agentic pipelines 18:16 Reviewing automated changes 25:08 Deployment and conclusion Local Workflows Huxley showcases how developers can use an MCP server to interact with Flyway Enterprise through local AI tools like Claude. • Developers can use the MCP to identify changes in the Chinook repository, edit schema models, and generate migration scripts without learning complex CLI commands. • The Flyway Enterprise engine acts as a deterministic gate, translating non-deterministic AI suggestions into safe, valid SQL based on established organizational rules. • This approach allows developers to work in preferred environments like VS Code or a terminal while ensuring all changes are version-controlled via the Flyway Enterprise project framework. Agentic Workflows The demo presents a future-ready, fully agentic workflow designed for enterprise automation through an autonomous copilot. • The agent follows instructions in an agent.md file to install Flyway Enterprise, set up a Docker container, and access necessary repositories. • Instead of writing raw scripts, the agent makes changes to a stateful schema model; subsequently, Flyway Enterprise generates the required migration and undo scripts. • A non-AI build stage deploys these scripts to a test database to verify syntax and impact before any human review occurs. • Automated code analysis via tools like CodeQL checks for rule violations such as SELECT * or TRUNCATE statements. • Unit testing is performed to ensure that business logic remains intact despite the automated changes. Governance Framework The framework ensures that security and governance are maintained even as development speed increases through AI integration. • By configuring policies within CI/CD platforms like GitHub, organizations can mandate branching policies and automated code reviews for both humans and AI agents. • The system prevents "hallucinated" code from reaching production by requiring successful builds and passing unit tests before alerting a human reviewer to a pull request. • This structure ensures that every change—regardless of its source—is subjected to the same rigorous deployment standards and drift checks against production environments. Learn more at redgate.com/flyway ✅ Subscribe to our YouTube channel to get notified on the latest video from us: https://youtube.com/@redgate For more updates follow us on: LinkedIn: https://www.linkedin.com/company/red-gate-software Website: https://www.red-gate.com/