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.

Dynamic Thinking Chain — DSH Plugin for DeepSeek Harness
DeepSeek Harness Plugin Hub
ProfilesPluginsCategoriesNewsDocsSign inManage Profiles
ProfilesPluginsCategoriesNewsDocsSign in
← Plugins
D

dynamic-thinking-chain

Dynamic Thinking Chain

DSH Dynamic Thinking Chain plugin: Automatically selects a reasoning level (off/low/high/max) for DeepSeek based on the difficulty of the current subtask. Each subtask split from a larger task and each sub-agent is assigned reasoning depth dynamically according to difficulty.

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

npx -y @deepseek-ai/dsh plugin --profile web add github:LovelyDazai/dsh-dynamic-thinking-chain#9fe495b5e00a7c8cae732fbb3ed364717fb76902
READMECompatibilityVersions

Compatibility and provenance

Dynamic Thinking Chain is published as dynamic-thinking-chain and currently resolves to version 0.1.2. The Hub verifies its manifest and preserves the exact installation source for reproducible installs.

DSH compatibility
*
Runtime surfaces
any
Release source
github
Registry updated
9/21/2026

Versions

0.1.2stable
9/21/2026

Related plugins

Loading related plugins…

Latest
0.1.2
DSH
*
HMR
Process restart
Tree shaking
Safe tree shaking not declared
Unpacked size
Unavailable
Files
Unavailable
Surface
any
License
MIT
Source
github
GitHub
★ 1
Weekly downloads
0
Last push
9/21/2026
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

Related plugins

More verified plugins in agents-orchestration.

Automation@michengai/dsh-automationDSH Automation — 在独立 DSH Session 中按计划执行和管理编码任务 · Schedule and manage coding tasks in isolated DSH sessionsHeadless@deepseek-ai/dsh-headlessThe dsh one-shot bundle: a direct core Agent/Session runner over dsh-base with no Host, HTTP, or browser layerAuto Reviewdsh-auto-reviewSecond-model AI auto-review for DeepSeek Harness approval requests: a read-only reviewer subagent decides allow/deny on the approval answerer chain, with fail-closed fallback and full session-log audit.Odai Dsh Agentodai-dsh-agentIndependently maintained odai governance mode preset that provides governance and collaboration as needed for tasks and can serve as the default overall controller.

README

dynamic-thinking-chain

让 DeepSeek 模型「按任务难度自动思考」的 DSH 插件。

DSH 会把一个大任务拆成很多小任务(todo_write 清单、子代理)。这个插件在每一次模型请求之前,根据当前正在做的那个小任务的难度,自动选择最合适的推理等级:

等级含义什么时候用
max最深思考规划/拆分大任务、算法、并发、安全、排查调试
high较深思考写函数、单元测试、改代码、写文档
low轻度思考打印 Hello World、改个名字这类琐碎小任务
off关闭思考默认禁用(DeepSeek 限制,见下方说明)

简单的活少想省 token,难的活深想保证质量——这就是「动态思维链」。

效果演示

一次真实运行里,一个大任务被拆成 4 个小任务,插件给每步分配了不同的推理等级:

[规划/拆分整个任务]    => max
[创建 hello.py]        => low
[二分查找 + 单元测试]  => high
[线程安全计数器]       => max
[写 README]            => high

安装

打包后用 dsh plugin 装进某个 profile:

cd dynamic-thinking-chain
npm pack   # 生成 dynamic-thinking-chain-0.1.2.tgz

dsh plugin --profile desktop add file:D:/workplace/dynamic-thinking-chain/dynamic-thinking-chain-0.1.2.tgz

装完重启对应界面:桌面版「退出后台运行」再打开;dsh web 重启 web 进程。

卸载:dsh plugin --profile desktop remove dynamic-thinking-chain。

配置(可选,默认开箱即用)

在 C:\Users\<你>\.dsh\profiles\<profile>\cordis.patch.yml 里覆盖:

- id: dynamic-thinking-chain
  config:
    enabled: true        # 总开关,设 false 彻底关闭
    lowThreshold: 0.30   # 难度分 < 0.30 → low
    highThreshold: 0.60  # 难度分 < 0.60 → high,否则 max
    planningMinLevel: high  # 拆分步的最低等级
    debug: false         # 设 true 才在控制台打印每次决策细节

常用配置项(完整列表见 src/classifier.js 的 DEFAULTS):

  • 打分:baseScore(基础分)、lengthSteps(文本长度阶梯)、hardWeight/hardCap、mediumWeight/mediumCap、easyWeight/easyCap、planningBias、errorBias/maxErrorBias、depthBias
  • 关键词:hardKeywords / mediumKeywords / easyKeywords(数组,可完全自定义;命中“简单”类词会降分)
  • 日志:debug(默认 false,静默;设 true 打印每次决策的难度分和理由)

插件默认不打日志。想看它每步怎么决策,把 debug: true 放进上面的 config;出错时才会打印一条 warning。

常见问题

1. 装了没效果? 插件只会对「支持推理等级」的模型生效。先确认当前会话选的模型是 DeepSeek(如 deepseek-v4-pro、deepseek-v4-flash);其他模型(比如 stepfun、普通对话模型)会被自动跳过、保持原样。

2. 为什么默认不用 off? DeepSeek 的规则:会话中途关闭思考会报 INVALID_REQUEST(上一轮的 reasoning_content 必须传回)。所以 offThreshold 默认是 0(永不开 off);即使你手动调大,插件也只会在会话第一条请求允许 off,之后自动退到 low。

3. 想固定某一档、不要动态? 把 enabled: false,或者在 DSH 界面右上角的模型选择里手动固定推理等级即可(本插件只在“动态”这件事上接管)。

原理(30 秒版)

  1. 插件监听每次模型请求前的 agent/request 瀑布,只改 reasoningEffort 一个字段。
  2. 难度来自当前会话信号:todo_write 里 in_progress 的小任务文本、最新的人类指令、工具报错次数、是否是第一步/子代理。
  3. 打分 → 映射到 off/low/high/max → 用 llm.resolveModelInfo 确认模型支持该等级 → 写入请求。
  4. 任何一步出错都原样放行,插件永远不会把请求弄坏。

开发

npm test   # 运行单元测试(打分、分级、等级回退)

插件本体零运行时依赖;profile 里有 @deepseek-ai/schemastery 和 @deepseek-ai/dsh-settings 时会自动注册 GUI 设置面板,否则自动跳过。

许可证

MIT