Hacker News
Canary (YC W26): AI QA That Reads Your Codebase, Generates Workflow Tests, Comments on PRs with Screen Recordings
Canary, from former Windsurf/Cognition/Google engineers, analyzes pull requests by reading the full codebase to understand routes, controllers, and validation logic, then generates and executes tests for affected user workflows — commenting directly on PRs with test results and screen recordings. Unlike general AI coding tools, it ships custom browser fleets, ephemeral environments, data seeding, and device farms for reliable execution. Internal QA-Bench v0 shows it outperforms GPT, Claude Code, and Sonnet on coverage metrics; HN commenters raised valid concerns about test stability post-merge and differentiation from existing code review tooling.
↳ Follow the thread