viego-qiin/dsh-loop-detector

Plugin插件 ⭐ 2 MIT Productivity & Tasks效率与任务

Dead-loop / repetition detector for DeepSeek Harness agents — hard-interrupts text loops and same-topic search retries that soft reminders miss.

Project Overview项目介绍

dsh-loop-detector is a repetition-detection plugin for DeepSeek Harness (dsh). It monitors the assistant stream in real time, catching text-level loops (period scan, 64-char n-gram, 512-char large anchor) and intent-level loops in web_search queries (keyword topic set, <25% novel tokens twice). On hit, it calls agent.steer() for guided retry or agent.cancel() to hard-abort the turn. Use it when local or quantized models ignore soft reminders and get stuck repeating output or re-searching the same topic. Note: detection is heuristic and per-session only; tune minLen and thresholds for your model.

dsh-loop-detector 是 DeepSeek Harness(dsh)的死循环与重复检测插件。它实时监测助手流,识别文本级重复(周期扫描、n-gram、512 字符大锚点)和 web_search 工具调用意图级重复(基于关键词主题集),命中后通过 agent.steer() 引导或 agent.cancel() 强制中断。适用于本地量化或推理模型在长输出或连续搜索时陷入同义循环而内置提醒无效的场景。注意:检测为启发式且仅在单会话内生效,需按模型调优 minLen 与阈值。

Or use CLI install (for developers)或使用命令行安装(适合开发者)

CLI Install命令行安装

dsh plugin --profile web add dsh-loop-detector

viego-qiin/dsh-loop-detector 加入你的 DSH 配置(web profile)即可启用。

READMEREADME

dsh-loop-detector

Dead-loop / repetition detector plugin for DeepSeek Harness (dsh).

npm License: MIT

Local LLMs (especially quantized / RL-tuned reasoning models) sometimes get stuck: repeating the same sentence forever, or re-running web_search with reworded queries for the same topic without gaining any new information. This plugin watches the agent's stream in real time, detects both kinds of loops, and interrupts them hard — not a soft reminder the model can ignore.

中文文档见 README.zh.md.

Features

  • Text-level repetition detection — watches assistant/chunk (text-delta + optional reasoning-delta) with three layered detectors:
    1. Period scan: tail repeats with a fixed period (8–256 chars)
    2. n-gram: a 64-char block seen ≥3 times in the recent window
    3. Large-scale anchor: a 512-char tail block reappearing earlier (catches whole-paragraph repetition, e.g. a model re-planning the same answer twice)
  • Tool-call intent-level repetition detection — watches tools/execute for web_search: extracts query keywords, keeps a "topic set" (union of the last 3 queries), and if a new query brings <25% novel tokens twice in a row, it judges the model is re-searching the same topic with different wording (intent-level loop, invisible to plain string matching).
  • Hard interrupt: on hit, agent.steer() injects a guidance message (auto retry, configurable); when retries are exhausted, agent.cancel() forcibly aborts the turn. Counters reset on turn/end.

Why hard interrupt? dsh's built-in repeat-tool-reminder only reminds the model. We observed local models (e.g. Qwen/Ornith 35B MoE with thinking off) ignore the reminder and keep looping — that's what this plugin exists for.

Quick start

Prereqs: a running dsh profile (see deepseek-harness docs).

Option A (recommended): install from npm (published to npm registry)

dsh plugin --profile web add dsh-loop-detector
dsh web   # restart to activate

Option B: install from a local copy

cp -r dsh-loop-detector $env:USERPROFILE\.dsh\profiles\web\node_modules\dsh-loop-detector

Then, in both cases, mount it via your profile's permanent patch layer ($env:USERPROFILE\.dsh\profiles\web\cordis.patch.yml):

- insert:
    - id: loop-detector
      name: 'dsh-loop-detector'
      config:
        minLen: 512          # only check output >= 512 chars (avoid short-text false positives)
        maxRetries: 1        # auto-retry 1x via steer; cancel when exhausted
        checkReasoning: true # also check reasoning-delta (thinking content)

Restart and verify:

dsh web

You should see the plugin loaded; when a loop is caught, the log shows:

[loop-detector] session xxx 检测到死循环(1/1): large-scale repeat: ...;steer 引导重试
[loop-detector] session xxx 连续 2 次搜索同一主题(无新信息): "...";steer 引导停止搜索

A ready-to-use cordis.patch.yml example is included in this repo.

Configuration

Field Default Meaning
minLen 512 Start text detection only after this many characters (avoid false positives on short replies)
maxRetries 1 Times to auto-retry via steer before hard cancel
checkReasoning true Also scan reasoning (thinking) content for repetition

How the detectors work

Text-level (detectRepetition)

1) period scan     tail 4 periods identical, period T in [8..256]        -> loop
2) n-gram          64-char block seen >= 3x in last 2048 chars           -> loop
3) large anchor    512-char tail block reappears earlier, extends >=512  -> loop

Tool-call intent-level

topic set = union of keywords of the last 3 web_search queries
new query  -> novel ratio = (# tokens not in topic set) / total tokens
novel ratio < 25% twice in a row  -> intent-level repetition -> steer/cancel

Verified against real cases: a Qwen3.5-9B stuck on an 82-char sentence (period=132 caught), a 2120-char duplicated planning block caught, and an 11-query same-topic search streak caught at query #8 (no false positive on the first 7 queries that each introduced new dimensions).

Scope & limitations

  • Intent-level detection currently targets web_search; other tools can be added by extending the exec.name filter.
  • The detectors are heuristic — tune minLen / thresholds for your model.
  • Detection is per-session and resets on turn/end (no cross-turn carryover).

License

MIT — see LICENSE.

上一个 Prev dsh-aurora-bg 下一个 Next dsh-notifier