DeepSeek Harness Plugin Hub

Publish and manage complete Harness Profiles. Discover Plugins for your next setup.

Explore

PluginsPresetsDocsNews

Community

Publish a pluginContactReport an issue

Resources

Plugin Hub on GitHubDeepSeek HarnessSystem statusPrivacy notice
© 2026 DeepSeek Harness Plugin HubPowered byPaxTech

Independent and unofficial. Not affiliated with, authorized by, or endorsed by DeepSeek.

Behavior Enhancer — DSH Plugin for DeepSeek Harness
← Plugins
B

dsh-behavior-enhancer

Behavior Enhancer

Behavior management plugin: system prompt behavior constraints / failure convergence parallelism / user intervention after consecutive failures, based on the real DSH plugin API

The plugin will be installed here. Keep web if you are unsure.

npx -y @deepseek-ai/dsh plugin --profile web add github:Liora2050348900/dsh-behavior-enhancer#e8a2e411135bb704f754babb1e7de0e6afaf9e7d
READMECompatibilityVersions

Compatibility and provenance

Behavior Enhancer is published as dsh-behavior-enhancer and currently resolves to version 1.0.0. The Hub verifies its manifest and preserves the exact installation source for reproducible installs.

DSH compatibility
*
Runtime surfaces
any
Release source
github
Registry updated
9/6/2026

Versions

1.0.0
stable
9/6/2026

Related plugins

Loading related plugins…

Latest
1.0.0
DSH
*
HMR
Process restart
Tree shaking
Safe tree shaking not declared
Unpacked size
Unavailable
Files
Unavailable
Surface
any
License
MIT
Source
github
Weekly downloads
0
View source ↗
README badge

Click the badge to copy Markdown for your README.

Do you maintain this Plugin?Claim benefit · Priority security scan

Verify the GitHub repository declared in package.json to manage this listing. After you claim it, Hub will prioritize a security scan of the current version and publish the result when it passes.

Claim this Plugin →
Report an issue
DeepSeek Harness Plugin Hub
ProfilesPluginsCategoriesNewsDocsSign inManage Profiles
ProfilesPluginsCategoriesNewsDocsSign in

Related plugins

More verified plugins in productivity-workflow.

Acp App@deepseek-ai/dsh-acp-appThe dsh ACP profile bundle: automation-only JSON-RPC stdio and process lifecycle over dsh-baseClient Ui Task Board@linxin666/dsh-client-ui-task-boardHost-authoritative task board for the DSH Web GUI with real session execution, Host cron scheduling, and optional cross-platform idle-sleep protection; mounted without DSH source changes.Web All@linxin666/dsh-web-allDSH Web UI 全家桶聚合插件:一键安装全部功能插件(task-board / git-graph / pet / remote-web-ui / web-ui-settings / skin-center / community-plugins / compat shim)。compat 桥接层已并入本包(src/client),无需独立 compat npm 包。Agent Teams@nanmicoder/dsh-agent-teamsAgentTeams for DeepSeek Harness: multi-agent team collaboration (captain, members, tasks with dependencies, messaging) driven by natural language, with a tree monitor in the web GUI

README

dsh-behavior-enhancer

DeepSeek Harness 行为管理插件:改变模型怎么调用工具,让复杂任务更稳定,不干预内容本身。 与 dsh-token-optimizer 职责互补—— token-optimizer 管内容(压缩/裁剪/采样),本插件管行为(调用纪律/失败收敛/连续失败介入)。两者互不依赖,可独立安装。

它做什么(30 秒版)

模块机制说明
behaviorPromptctx.inject(['systemPrompt']) + systemPrompt.section()注册静态行为约束段(先读后写、批量前先验证、失败即收敛、同类修改串行)。软约束,强制靠下面模块
parallelConvergencetools/result + settings.update('agent-loop', {maxParallelToolCalls})任何工具失败(含命令级失败)→ 并行池上限压到 1;连续 3 次成功 → 恢复用户原值;上下文压力 ≥60%(idle 测量)→ 降档,回落 <54% 滞回恢复。串行状态跨进程持久化,重启自动接管还原
failureGuardtools/result + agent.followup()同工具连续失败 ≥2 次 → 向模型注入提示,引导它向用户确认而不是继续重试

失败识别两档:工具级 isError + 命令级签名(stderrAsFailure,默认开)——DSH 的 pwsh 工具只有基础设施错误才标 isError,命令失败(非零退出/stderr 错误)作为普通结果返回;本插件按 [exit code: N] 与 stderr 错误关键词补识别,普通 warning 不误判。

设计依据(DSH 核心源码核实)

  • 核心已内置写操作排他:isConcurrencySafe 分类(read/read_image 并行,edit/write/pwsh 等 exclusive 屏障),插件不需要做 per-tool 串行规则;
  • 插件唯一动态并行杠杆是池上限 maxParallelToolCalls(settings 活 getter,运行时 update 即生效);
  • "暂停其他进行中的调用"无 API,不支持;工具失败无核心重试,改为"降并行让模型逐步修正";
  • 行为约束段是静态文本,保护 DeepSeek 前缀缓存(实测命中率 99.75%,动态内容会杀缓存)。

安装

dsh plugin --profile web add ./dsh-behavior-enhancer

配置

- id: behavior-enhancer
  name: 'dsh-behavior-enhancer'
  config:
    behaviorPrompt:
      enabled: true
      order: -98
      text: |        # 可整体替换;必须保持静态(禁时间戳/动态值)
        ## 工具调用行为纪律(本会话)
        - 先读后写
        - 批量操作前先验证
        - 失败立即收敛
        - 同类修改串行
        - 证据与诚实
        - 上下文已压缩时先查日志再回答
    parallelConvergence:
      enabled: true
      recoveryThreshold: 3     # 连续成功次数达到后恢复原并行度
      complexRatio: 0.6        # 上下文压力比(idle 测量),超阈值降档
      minParallel: 1
      defaultParallel: 10      # 读不到当前值时的 fallback(核心默认 10)
      contextWindowFallback: 1000000
    failureGuard:
      enabled: true
      maxFailures: 2
      followupMessage: '工具 {tool} 已连续失败 {count} 次。请停止重试,向用户说明失败原因并询问下一步。'
      blockFeedback: false     # 慎用:拦截错误结果正文,模型会看不到失败细节
      feedbackText: '该工具调用失败。请先分析失败原因并修正调用参数后再重试,不要原样重复。'

开发

node test/smoke.mjs          # 单进程全量自检(无需 API),npm test 同

路线图

  • 实验分支:给 grep/glob 补 isConcurrencySafe(需 agent 作用域 shadow 包装器,复杂度高,主线不依赖)
  • 硬阻塞选项:连续失败后用 userQuestions.ask 阻塞等用户回答(默认走软通道 followup)