Agree on verifiable completion criteria before writing code with agents
Larsen Cundric proposes a rule for writing code with agents: before requesting a change, first agree on how both sides will confirm the change works. His acceptance methods include a test that currently fails, a command you can run yourself, and logs or screenshots of real run results; a "done" without evidence will be sent back. He believes most failed agent sessions start with vague requirements and missing checks, and he asks readers to share their own criteria for trusting an agent's "done."
Why read this
AI score 57
The author puts acceptance criteria up front as executable checks, so readers can use them to reduce rework caused by vague deliverables in agent coding tasks.
Topics
Body
Rule I'd give anyone coding with agents:
Before you ask for the change, ask how you'll both know it worked.
> a test that fails right now > a command you can run yourself > a log or screenshot of the real thing > "done" with no proof gets sent back
Most bad agent sessions start with a vague ask and no check.
What's your rule for trusting an agent's "done"?