Tool-loop detection

기준일: 2026-07-26
난이도: 중급
공식 기준: Tool-loop detection

개요

이 페이지는 OpenClaw Tool-loop detection 도구(파라미터, 권한, 설정, CLI)를 공식 문서 기준으로 정리합니다.

공식 요약: How to enable guardrails that detect repetitive tool-call loops

도구 가시성은 profile / allow·deny policy / sandbox / channel 권한에 따라 달라집니다. 최신 스키마는 항상 공식 문서를 우선합니다.

공식 문서 기반 상세

아래는 공식 tools/loop-detection 문서를 정리한 내용입니다. 코드 블록, 파라미터 이름, 기본값은 원문 그대로입니다.

OpenClaw has two cooperating guardrails against repetitive tool-call patterns, both configured under tools.loopDetection:

1. Loop detection (enabled) - disabled by default. Watches the rolling tool-call history for repeated patterns and unknown-tool retries.
2. Post-compaction guard - enabled whenever enabled is not explicitly false. Arms after every compaction-retry and aborts the run if the agent repeats the same (tool, args, result) triple within the window.

Set tools.loopDetection.enabled: false to silence both guardrails.

Why this exists

• Detect repetitive sequences that make no progress.
• Detect high-frequency no-result loops (same tool, same inputs, repeated errors).
• Detect specific repeated-call patterns for known polling tools.
• Break context-overflow -> compaction -> same-loop cycles instead of letting them run indefinitely.

Configuration block

Global setting:

{
  tools: {
    loopDetection: {
      enabled: false, // master switch for the rolling-history detectors
    },
  },
}

Per-agent override (optional, at agents.entries.*.tools.loopDetection):

{
  agents: {
    list: [
      {
        id: "safe-runner",
        tools: {
          loopDetection: {
            enabled: true,
          },
        },
      },
    ],
  },
}

The per-agent setting overrides the global setting.

Field behavior

Field	Default	Effect
enabled	false	Master switch for the rolling-history detectors. false also disables the post-compaction guard.

For exec, no-progress hashing compares stable command outcomes (status, exit code, timed-out flag, output) and ignores volatile runtime metadata such as duration, PID, session ID, and working directory. Outbound message-send results are hashed with volatile per-call ids (message id, file id, timestamp) stripped, so a "sent" result does not look identical to a different "sent" result. When a run id is available, history is evaluated only within that run, so scheduled heartbeat cycles and fresh runs do not inherit stale loop counts from earlier runs.

Recommended setup

• For smaller models, set enabled: true. Flagship models rarely need rolling-history detection and can leave the master switch false while still benefiting from the post-compaction guard.
• To disable everything, including the post-compaction guard, set tools.loopDetection.enabled: false explicitly.

Post-compaction guard

After a compaction-retry following a context-overflow, the runner arms a short-window guard on the next few tool calls. If the agent emits the same (toolName, argsHash, resultHash) triple enough times within that window, the guard concludes compaction did not break the loop and aborts the run with a compaction_loop_persisted error.

The guard is gated by the master tools.loopDetection.enabled flag with one twist: it stays enabled when the flag is unset or true, and only turns off when the flag is explicitly false. This is intentional - the guard exists to escape compaction loops that would otherwise burn unbounded tokens, so a no-config user still gets the protection.

{
  tools: {
    loopDetection: {
      // master switch; set false to disable the guard along with the rolling detectors
      enabled: true,
    },
  },
}

• The guard never aborts while results are changing; only byte-identical results across the window trigger it.
• It only arms in the immediate aftermath of a compaction-retry, not at other points in a run.

The post-compaction guard runs whenever the master flag is not explicitly false, even if you never wrote a tools.loopDetection block. To verify, look for post-compaction guard armed for N attempts in the gateway log immediately after a compaction event.

Logs and expected behavior

When a loop is detected, OpenClaw logs a loop event and either warns or blocks the next tool-cycle depending on severity, protecting against runaway token spend and lockups while preserving normal tool access.

• Warnings come first.
• Blocking follows once a pattern persists past the warning threshold.
• Critical thresholds block the next tool-cycle and surface a clear loop-detection reason in the run record.
• The post-compaction guard emits compaction_loop_persisted errors naming the offending tool and identical-call count.

관련 문서

Allow/deny policy for shell execution.


Reasoning effort levels and provider-policy interaction.


Spawning isolated agents to bound runaway behavior.


Full `tools.loopDetection` schema and merging semantics.

검증 체크리스트

• 해당 tool이 활성 profile/policy에서 허용되는지 확인
• sandbox / elevated / host 실행 경로 정책을 이해했는지 확인
• 채널·에이전트 권한과 충돌하지 않는지 확인
• 공식 CLI/--help와 문서 옵션이 버전과 맞는지 확인

관련

• Tools 목록 인덱스
• 공식: Tools overview, Tools and custom providers