Agents should not grade their own work
can fail when each step decides by itself that its own work is good enough. If the first agent does research, marks it complete, and passes it on, the next agent may trust that mark and build on a mistake. By the time the final agent returns an answer, every gate may show success even though nobody independently checked the early work.
This is : the system can look healthy in the logs while producing a wrong result. The check is to mark every point where the producer of a result also decides whether it moves forward. Prompt instructions that let an agent continue based on its own sense of completeness or confidence are warning signs.
The fix is structural, not just a stronger prompt. The agent that produces the output should be separate from the role that judges it.
Key points
- A pass mark from an agent is weak evidence if that same agent created the work.
- Successful logs can hide bad early .
- Audit prompts for gates where an agent moves forward based on its own completeness or confidence check.
- Separate the producer from the judge at important gates.