| name | loop-detector |
| description | Use when commands, searches, file reads, edits, tests, critics, plans, context reloads, or strategies repeat without new evidence or acceptance progress. |
Loop Detector
Core principle
Detect semantic repetition, not only identical syntax; high activity plus flat outcome is probable livelock.
Procedure
- List recent actions and their intended evidence.
- Group semantically equivalent actions.
- Compare completed requirements, removed blockers, and newly passing tests.
- Mark repeated actions that added no information.
- Identify retry, tool, edit-revert, critic, replan, context, or oscillation patterns.
- Trigger attention or strategy recovery at the configured threshold.
Required output
Loop type, repeated strategy, unchanged outcome, threshold, and recovery action.
Limits
Same strategy 3; critics 2; context reloads 2; stalled cycles 4.
Common mistakes
Do not call renamed commands or alternate tools meaningful novelty when they test the same hypothesis.
Stop condition
Stop when repetition is interrupted and a materially different bounded action is selected.
