Tests that write and repair themselves
Stably AI was founded in San Francisco by Jinjing Liang, who built Chrome's testing and release infrastructure at Google, and Neil Parker, one of Uber's youngest tech leads. Backed by Y Combinator, the company attacks the most hated chore in software engineering: writing and maintaining end-to-end tests.
How it works
Developers describe test scenarios in plain English and Stably converts them into real Playwright code — no vendor lock-in, since the output is standard open-source test scripts. The platform's self-healing engine automatically updates tests as the product's UI and flows evolve, eliminating the maintenance burden that causes most test suites to rot. Stably runs in CI on every pull request and claims teams can reach 80% test coverage in minutes rather than months, with three access paths: a CLI, a cloud platform, and a visual editor. Integrations span Jira, git repositories, documentation systems, and CI/CD pipelines.
Orca: open-source agent orchestration
Stably also builds Orca, a free MIT-licensed terminal IDE that orchestrates multiple AI coding agents — Claude Code, Codex, Gemini, and Cursor CLI — working in parallel inside isolated git worktrees. Paired with Stably's testing platform, it forms a build-and-verify loop for agentic development, with paid plans starting around $39/month.