The idea

CodeRabbit CEO Harjot Gill describes using agents to ship software quickly, while maintaining quality through layered verification, staged rollouts, and human review of releases rather than individual pull requests.

Methods

  1. Use agents for recurring engineering work such as fixing flaky tests, improving CI, and refining caching.
  2. Give agents ways to verify their work: use simulations, API emulation, development environments, and staged rollouts.
  3. Review plans and releases that combine multiple pull requests, rather than reviewing every agent-generated pull request individually.

Try this today

Choose one recurring engineering task and have an agent tackle it with a clear way to verify the result.

We'll check in tomorrow to ask whether it worked. Your answer stays in this browser; we only count totals.

Go deeper with the original

These takeaways are our summary. The full argument, evidence and voice are in Gergely Orosz's piece.

Read the original on The Pragmatic Engineer (opens in a new tab)