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.

Model Think Level — DSH Plugin for DeepSeek Harness
DeepSeek Harness Plugin Hub
ProfilesPluginsCategoriesNewsDocsSign inManage Profiles
ProfilesPluginsCategoriesNewsDocsSign in
← Plugins

dsh-model-think-level

Model Think Level

Adds thinking levels for DSH’s third-party models: injects a Thinking toggle and intensity/modality editor into each model row on the official Models page, adds a provider → model selector, visual badges, and level rows to the input model menu, with search, drag-and-drop sorting, and provider enable

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

npx -y @deepseek-ai/dsh plugin --profile web add dsh-model-think-level@1.1.5
READMECompatibilityVersions

Description

Adds thinking levels for DSH’s third-party models: injects a Thinking toggle and intensity/modality editor into each model row on the official Models page, adds a provider → model selector, visual badges, and level rows to the input model menu, with search, drag-and-drop sorting, and provider enable/disable switches; also takes over request headers and user-agent by route.

Compatibility and provenance

Model Think Level is published as dsh-model-think-level and currently resolves to version 1.1.5. The Hub verifies its manifest and preserves the exact installation source for reproducible installs.

DSH compatibility
*
Runtime surfaces
web
Release source
npm
Registry updated
10/16/2026

Versions

1.1.5stable
10/16/2026
1.1.4stable
10/16/2026
1.1.3stable
10/15/2026
Show 11 more versionsCollapse versions
1.1.2stable
10/15/2026
1.1.1stable
10/13/2026
1.1.0stable
10/11/2026
1.0.7stable
10/10/2026
1.0.6stable
10/9/2026
1.0.5stable
10/9/2026
1.0.4stable
10/8/2026
1.0.3stable
10/8/2026
1.0.2stable
10/7/2026
1.0.1stable
10/7/2026
1.0.0stable
10/7/2026

Related plugins

Loading related plugins…

Latest
1.1.5
DSH
*
HMR
Process restart
Tree shaking
Safe tree shaking not declared
Unpacked size
3.4 MB
Files
79
Surface
web
License
MIT
Source
npm
GitHub
★ 3
Weekly downloads
1,724
Security scan
✓ v1.1.5 scan passed
Last push
10/16/2026
View source ↗Project homepage ↗
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.

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.Codex Subscriptiondsh-codex-subscriptionUse ChatGPT and Codex subscriptions in DeepSeek Harness with OAuth, quota, safe resets, web search, images, and Fast modePlugin@opencode2dsh/dsh-pluginFree OpenCode Zen models for DeepSeek Harness (DSH): native dsh-llm adapter plugin, no API key.

README

DSH Model Think Level

DSH 的第三方模型补齐思考等级:在官方 Models 页每个模型行注入 Thinking 开关和强度/模态编辑器,在输入框模型菜单加供应商→模型选择器、视觉徽章与等级行,并带搜索、拖拽排序、供应商启用开关;以及按路由接管请求头与 user-agent。

模型选择器

供应商与模型选择器 思考等级与模型搜索

供应商配置

供应商配置面板 模型能力配置 供应商高级配置

兼容字段与已有配置

1.1.4 起不再提供 thinkingTokenBudgetField、vllmPriority、supportsMaxOutputTokens 的专用编辑控件。已有值仍由 DSH 的 pi-ai 适配器使用;修改模型的思考档位或输入模态时,插件会保留这些值和其他手动兼容配置,避免覆盖供应商的定制行为。

需要调整旧值时,先备份当前 profile 的 cordis.patch.yml(位于 DSH 数据目录的 profiles/<profile>/),找到 llm-pi-ai 对应供应商、模型行的 compat,仅修改或删除所需字段,保存后重新加载该 profile:

  • thinkingTokenBudgetField:令牌预算的请求字段名,可设为 thinking_token_budget、thinking_budget 或 thinking_budget_tokens。删除后使用默认字段;启用令牌预算的模型在后续适配或保存时可能重新补入 thinking_token_budget。若端点不支持令牌预算,应将 supportsThinkingTokenBudget 设为 false。
  • vllmPriority:删除后不再发送定制的 vLLM 优先级。
  • supportsMaxOutputTokens:删除后恢复适配器默认行为;保留 false 会继续关闭对应输出令牌参数。

不要删除整个供应商配置,也不要将包含 API key 的配置文件贴到公开 issue。

开发与验证

使用 Node.js ^22.19.0 || >=24。依次运行 npm ci、npm run typecheck、npm test、npm run build、npm run check:package。测试最多并行 4 个 worker,以控制 DOM 测试的内存占用;仍使用 Vitest 默认超时。

check:package 检查 npm 实际打包清单中的入口、截图和声明文件依赖,CI 与发布流程都会执行。类型声明所引用的 DSH 可选 peer 由宿主开发环境提供,不会自动安装额外宿主插件。

更新模型知识表后,可运行 npm run docs:models 生成可查阅的模型列表。node scripts/wire-repro.mjs 使用当前安装的 pi-ai,向本地回显服务发送测试请求,无需供应商凭据。