Definition
Plain language
A safety layer that watches for an AI agent looping or going nowhere and forces it to rethink.
As stated in the literature
In Life-Harness, the runtime layer monitoring recent action history for repetition or lack of progress and injecting recovery prompts when trajectory degeneration is detected.
Why it matters: It rescues agents from common stuck states that would otherwise turn a minor confusion into a complete task failure.
For example, if an agent has clicked the same button five times in a row, trajectory regulation injects a prompt nudging it to try a different approach.
Heard on the show
“The trajectory regulation layer watches the agent's recent actions, notices when it's repeating itself or burning steps without progress, and injects a recovery prompt.”Episode 071 — When the Model Is Fine and the Plumbing Is Broken: Fixing Agents at the Interface