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.

DeepSeek Harness Plugin Hub
ProfilesPluginsCategoriesNewsDocsSign inManage Profiles
ProfilesPluginsCategoriesNewsDocsSign in
← Plugins
R

@zhourenke/dsh-reasoning-mode

Reasoning Mode

Configure Responses reasoning mode and summary per provider/model in DeepSeek Harness

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

npx -y @deepseek-ai/dsh plugin --profile web add github:zhourenke/dsh-reasoning-mode#e06fb90a21fce03b1e5f770c996d2762cdfd4541
READMECompatibilityVersions

Compatibility and provenance

Reasoning Mode is published as @zhourenke/dsh-reasoning-mode and currently resolves to version 0.2.0-rc2.2. The Hub verifies its manifest and preserves the exact installation source for reproducible installs.

DSH compatibility
*
Runtime surfaces
web
Release source
github
Registry updated
10/18/2026

Versions

0.2.0-rc2.2prerelease
10/18/2026
0.1.0stable
9/14/2026
Show 1 more versionCollapse versions
0.2.0-rc2.1prerelease
9/30/2026

Related plugins

Loading related plugins…

Latest
0.2.0-rc2.2
DSH
*
HMR
Process restart
Tree shaking
Safe tree shaking not declared
Unpacked size
Unavailable
Files
Unavailable
Surface
web
License
MIT
Source
github
GitHub
★ 1
Weekly downloads
0
Last push
10/18/2026
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 models-usage.

Experimental Cot Translation Bundle@deepseek-ai/dsh-experimental-cot-translation-bundleExperimental machine translation of expanded reasoningOur Free Model@ebony-vinyl/dsh-our-free-modelOur Free Model — Install this plugin in dsh to use cutting-edge models including DeepSeek V4.1 Flash and Kimi K3, with no login, registration, or API Key required. The desktop version also offers an EAC co-payment channel; starting with 2.0, it supports ten account channels: LobsterAI, Qoder, TRAE, Usage@linxin666/dsh-usageUsage statistics plugin for the dsh web GUI: per-provider balance and coding-plan quota detection plus a live token usage ledger, with a compact price glance card above the sidebar's Settings rowClient Ui Model Capabilities@linxin666/dsh-client-ui-model-capabilitiesPer-model capability declarations (image input and reasoning efforts) for custom pi-ai providers, edited in place on the Models settings cards; writes the official llm-pi-ai settings namespace, no DSH source changes.

README

English | 中文

@zhourenke/dsh-reasoning-mode

给选定的 provider/model 指定 Responses 请求的推理模式与摘要等级。

DSH 自带 adapter 提供的 reasoningEffort(每个模型的推理档位),但 Responses 请求体里的 reasoning.mode(Standard / Pro)与 reasoning.summary(Auto / Concise / Detailed)没有逐路由入口;本插件补上这一层:只改写你勾选的路由,其它请求原样通过。

它解决什么问题

  • 逐路由控制推理强度:同一个 model 在不同 provider 下互不影响。
  • 摘要等级可调:Auto 交给模型自己判断,Concise / Detailed 强制要短或要长。
  • 随时可卸:作为 profile 层插入,不修改 DSH 本体;清空勾选即完全关闭。

前置条件

走 OpenAI Responses 协议的 GPT 模型。走 Chat Completions 的模型或其它不支持该配置项的模型不会生效。

安装

dsh plugin --profile web add "github:zhourenke/dsh-reasoning-mode"

安装后需要重启 DSH 并刷新页面(宿主半边在进程启动时装载,浏览器半边在插件激活时取一次产物),然后在左侧工作区顶部的 插件 选项卡里找到 @zhourenke/dsh-reasoning-mode,在它页面的 包含的组件 列表中点 reasoning-mode 那一行的 配置,即可打开本插件的配置页。(0.2.0 起,插件的配置页只在这里;设置页里那一项已改名 内置插件,只列本部署自带的插件,没有本插件的入口。)

卸载:

dsh plugin --profile web remove @zhourenke/dsh-reasoning-mode

快速上手

  1. 打开左侧工作区顶部的 插件 选项卡,进入 @zhourenke/dsh-reasoning-mode,在 包含的组件 里点 reasoning-mode 行的 配置,在模型目录里勾选要让插件改写的 provider/model,点保存。
  2. 打开一个会话,在模型选择器中把模型切到刚才勾选的路由。
  3. 点开输入栏右侧出现的标签:卡片上半是 推理模式(Standard / Pro),下半是 摘要等级(Auto / Concise / Detailed);当前取值带勾,点一下就生效(没有保存按钮)。

新勾选的路由默认 Standard + Auto;取消勾选再重新勾选会保留原来的值。

配置保存即生效,不需要重启:本插件的配置存在 profile 补丁层里,而补丁层是热重载的。

配置

配置的落点是 web profile 的补丁层 <harness home>/profiles/web/cordis.patch.yml,挂在 reasoning-mode 这一行下面:

- id: reasoning-mode
  name: '@zhourenke/dsh-reasoning-mode'
  config:
    models:
      - provider: <provider-id>
        model: <model-id>
        mode: pro
        summary: detailed
字段类型默认说明
modelsarray[]要改写的路由列表,每项为 { provider, model, mode, summary }。空列表是唯一的关闭状态。
models[].providerstring无provider id,必须与目录里的 id 逐字一致。
models[].modelstring无model id,同样逐字一致。
models[].modestandard | prostandard下发到 reasoning.mode。
models[].summaryauto | concise | detailedauto下发到 reasoning.summary。
  • provider 与 model 必须同时匹配;同一个 model 挂在另一个 provider 下不会命中那条路由。
  • 缺失或写错的 mode / summary 会在读取时归一化回 standard / auto,不会导致加载失败;完全相同的 provider+model 重复条目会被去重。
  • id 写错的条目不会报错,只是不生效(在配置页里勾选就不会写错)。
  • 目录里当前不存在的已保存路由会保留在「已保存但当前不可用」分组里,手动取消勾选并保存后会消失。
  • 插件不判断模型是否支持推理:勾选一个不接受 reasoning 的模型时,上游会直接返回「不支持该参数」的报错(实测)。这是有意为之——报错比静默忽略更容易定位。

兼容性

在 v0.2.0-rc.2(2026-09)下测试通过。 在 v0.2.1-rc.2(2026-10)下测试通过。

许可证

MIT