Kerno v1 launches as an AI runtime QA engineer for backend code

Summary

Kerno has released v1 of its AI runtime QA product for backend code. It works inside Claude Code, Cursor, Codex and other coding agents that support MCP, testing critical user flows and backend entry points against real databases, queues and services before changes merge. Kerno first indexes a repository, maps HTTP routes, MCP tools, background jobs and queue consumers, then helps an agent define flows, generate baseline tests and rerun tests when code changes affect them. The release adds end-to-end flow testing, configurable repository rules, and security checks covering applicable OWASP API and Web Top 10 categories, with evidence attached to failures. It supports workers and queues including Celery, Kafka, RabbitMQ, SQS, Azure Service Bus and Graphile Worker, as well as MCP servers and AWS serverless applications using SAM or Serverless Framework configurations. Tests are saved in the repository and can run through a deterministic GitHub Action on every pull request, with JUnit results. Kerno also introduces a Portal with analytics, coverage tracking, run reports, repository views and a Behaviour Drift heatmap for monitoring changes across teams. The company reports that teams using Kerno saw 6x faster merges for tested pull requests, 48% more runtime issues caught and 10 hours freed per engineer each week, but the article does not describe an independent evaluation. Setup takes about 30 to 60 minutes and requires Docker, Node.js 18 or later and a Git repository. The Community plan covers one repository and 30 monthly test runs without a credit card; Pro is listed at $50 per developer per month, while Enterprise adds SSO, self-hosting, audit logs and an SLA.