Budget review and testing time for agent tasks
A year with AI agents: the review and testing time adds up · X · @Gegam245074 · Sep 2026
A developer's takeaway after a year: agents don't automatically speed you up, because your review and testing time sits on top of the agent's run time.
The honest numbers
- Says about 40 min of agent time per small task
- Says about 30 min reviewing and about as long testing, around 2 hours total
As stated by the source. Not independently verified.
Lesson
Budget time for reading and testing what the agent wrote. Keep tasks small enough that you can review them in minutes.
Fix prompt
Copy this into your own AI tool (Cursor, Lovable, Claude Code, Bolt…). It's general guidance written by us. Review what the AI suggests before applying it.
my AI coding tool just finished a change in my app. Before I accept it, act as a careful reviewer. 1) Summarise in plain English what changed and why, file by file. 2) Point out anything risky: auth, payments, database queries, deleted code, new dependencies, or secrets. 3) List edge cases that might break (empty data, slow network, two devices, a logged-out user). 4) Give me a 10-minute manual test plan for the real user flows, and the automated tests worth adding. Don't make further changes until I reply.
Source: X ↗