DeepSeek Harness Plugin Hub

发布与管理完整 Harness Profiles,发现适合你的插件。

探索

插件目录环境预设文档中心动态

社区

发布插件联系我们报告问题

相关链接

Plugin Hub GitHubDeepSeek Harness 官方项目系统状态隐私说明
© 2026 DeepSeek Harness Plugin HubPowered byPaxTech

独立、非官方社区项目,与 DeepSeek 官方无隶属、授权或背书关系。

Keenable — DeepSeek Harness 插件(DSH Plugin)
DeepSeek Harness Plugin Hub
ProfilesPlugins分类动态文档登录管理 Profiles
ProfilesPlugins分类动态文档登录
← Plugins
K

@keenable/dsh-keenable

Keenable

Keenable 为 DeepSeek Harness 提供网页搜索和页面抓取功能:内置 web_search 和 web_fetch 工具,无需 API 密钥

插件会安装到这里;不确定时保持 web。

npx -y @deepseek-ai/dsh plugin --profile web add github:keenableai/dsh-keenable#b6f88dabe5d46c6130e31a0346bc6c7428d928ec
README兼容性版本

兼容性与来源证明

Keenable 以 @keenable/dsh-keenable 发布,当前版本为 0.1.0。Plugin Hub 会校验它的 manifest,并保存精确安装来源,便于复现安装结果。

DSH 兼容范围
*
运行环境
any
发布来源
github
Registry 更新时间
2026/9/29

版本

0.1.0stable
2026/9/29

相关插件

正在加载相关插件…

最新版
0.1.0
DSH
*
HMR
重启进程
Tree shaking
已声明 sideEffects: false
解包体积
未提供
文件数
未提供
Surface
any
许可证
MIT
发布源
github
GitHub
★ 0
周下载
0
最近提交
2026/10/13
查看源码 ↗
README Badge

点击下方 Badge 复制 Markdown,粘贴到 README 即可。

这是你的 Plugin?认领权益 · 优先安全扫描

验证 package.json 声明的 GitHub 仓库,即可管理这个公开页面。认领后,Hub 会优先安排当前版本的安全扫描,并在通过后公开展示结果。

认领这个 Plugin →
报告问题

相关插件

继续浏览 search-research 分类下经过校验的插件。

Weknora@wxg-prc-cpg/dsh-weknora适用于 DeepSeek Harness (dsh) 的 WeKnora 知识检索工具:通过自有知识库进行语义搜索、文档阅读以及 RAG/代理回答。Free Searchdsh-free-searchDeepSeek Harness 的免费网页搜索:24 个引擎(Bing/DuckDuckGo/AnySearch/SearXNG/Exa/Tavily/Keenable/Firecrawl/Parallel 无需密钥;通过 MCP 使用 Zhihu 全局/站内搜索,以及 Perplexity/SerpBase/DeepSeek/You.com/Baidu/Kimi/Aliyun/Doubao,需密钥;OpenAI/Gemini/ClaudeDejadsh-dejaDeepSeek Harness 的 deja-vu 记忆:其他 40 个编码代理的会话历史,可搜索,并在每个步骤前调用——本地索引,无需 LLM。Builtin Browserdsh-builtin-browser适用于 DeepSeek Harness 的共享真实浏览器插件:安装即用——可由人类接管的可见原生浏览器,并由代理通过 CDP 驱动。DOM 级交互(兼容 React/Vue)、按任务隔离会话、Cookie 持久化(dsh_browser_auth)、C

README

Keenable for DeepSeek Harness

Web search and page reading for DeepSeek Harness through Keenable, with no API key and no account. It works with whatever model you run in Harness, local models included.

The plugin points Harness's built-in web_search and web_fetch tools at Keenable. It adds no new tools: the model asks for a query or a URL exactly as before, and Keenable answers.

Why

The search provider Harness ships with needs a DeepSeek account or a DeepSeek API key. With a local or third-party model and neither of those, web_search fails. Keenable's public endpoints need no key, so search works on a fresh install:

  • web_search returns ranked web results with an excerpt of each page and its publication date, when known.
  • web_fetch returns a page as markdown. The page is retrieved on Keenable's servers, so a URL the model chose is never requested from the machine running Harness, and private or internal hosts are refused.

An API key is optional. It only lifts the rate limit described below.

Install

Requires DeepSeek Harness 0.1.7-rc.2 or later and Node.js 22.12 or later.

From Harness: open Plugins → Add plugin, enter @keenable/dsh-keenable, then Install → Enable now.

From the terminal: install into the profile you use, then restart Harness:

npx @deepseek-ai/dsh plugin --profile web add @keenable/dsh-keenable
npx @deepseek-ai/dsh web

Profiles are separate: web, tui and headless each need their own install. Replace web with the profile you run.

Then ask your model to search the web. No further setup is needed.

Limits and the optional API key

Without a key, requests go to Keenable's public endpoints, limited to 10 requests per second and 1,000 per hour per IP address. Search and fetch are counted separately. Everyone behind the same network address shares the limit, so a busy office or CI runner can reach it; the tool error then says so.

A key lifts the limit. Create one in the Keenable console, then either:

  • set KEENABLE_API_KEY in the environment Harness starts in, or
  • enter it as apiKey in this plugin's settings on the Plugins page.

A key in the settings wins over the environment variable.

Settings

SettingDefaultMeaning
apiKey$KEENABLE_API_KEYOptional. Switches both tools to the authenticated endpoints.
maxSnippetChars500Excerpt length per search result, 180 to 10,000 characters.
fetchLivetrueFetch pages live from the source. When off, web_fetch returns Keenable's indexed copy and fails for pages it has not indexed.
maxBodyChars100000Maximum characters of page text web_fetch returns. Longer pages are cut and marked as truncated.
baseURLhttps://api.keenable.aiAPI origin. HTTPS only.

Keeping the local fetch provider

The plugin pins both capabilities to Keenable. To keep Harness's local HTTP fetch provider and use Keenable for search only, override the web row in your profile's cordis.patch.yml:

- id: web
  config:
    searchProvider: keenable
    fetchProvider: http

Both fields are needed: a config patch replaces the row's whole config object.

Privacy

Search queries and fetched URLs are sent to Keenable, even when your model runs locally. Requests identify the plugin with an X-Keenable-Title: dsh-keenable header, which the public endpoints require. Results are external content, and Harness marks them as untrusted data for the model.

Troubleshooting

SymptomWhat to check
WEB_PROVIDER_CONFIGURED_MISSINGThe plugin is not loaded in this profile. Install it into the profile you run, and restart Harness.
"requests share a per-IP limit"The keyless limit was reached. Wait, or set an API key.
"Fetching from private/internal hosts is not allowed"Keenable does not fetch localhost or private addresses. Use the local HTTP fetch provider for those (see above).
Search still asks for DeepSeek credentialsAnother patch layer pins searchProvider back. Check the composed profile with npx @deepseek-ai/dsh --profile web --dump-config.

Development

npm ci
npm test            # offline: providers, plugin registration, model-facing tools, manifest
npm run test:live   # real Keenable API, no key

The tests load the plugin into a real Cordis context with Harness's own dsh-web seam and web_search/web_fetch tools. The code is plain JavaScript and needs no build step.

License

MIT