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
A

dsh-ai-relay

Ai Relay

Ask other AI models to help with the work: OpenAI-compatible and Anthropic routes as agent tools inside DeepSeek Harness.

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

npx -y @deepseek-ai/dsh plugin --profile web add github:pairun325/dsh-ai-relay#9cf3262e6aa180ebc3599ca22300ec7838f70d61
READMECompatibilityVersions

Compatibility and provenance

Ai Relay is published as dsh-ai-relay and currently resolves to version 0.1.0. The Hub verifies its manifest and preserves the exact installation source for reproducible installs.

DSH compatibility
*
Runtime surfaces
any
Release source
github
Registry updated
10/24/2026

Versions

0.1.0stable
10/24/2026

Related plugins

Loading related plugins…

Latest
0.1.0
DSH
*
HMR
Process restart
Tree shaking
Safe tree shaking not declared
Unpacked size
Unavailable
Files
Unavailable
Surface
any
License
MIT
Source
github
GitHub
★ 0
Weekly downloads
0
Last push
10/24/2026
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 rowConnect Workbuddydsh-connect-workbuddyConnects locally signed-in WorkBuddy models (domestic and international versions) to DeepSeek Harness, automatically reading each model’s points multiplier from the upstream catalog on every refresh, with per-model image input toggles, multi-account switching, automatic account rotation from an acco

README

dsh-ai-relay

让鲸鲸把活分给别的 AI:把 OpenAI 兼容的端点、Anthropic 的 Messages API、或者任何自建网关,变成 DSH 里的两个工具。

  • ai_ask — 把一段自包含的提问发给外部模型,拿回答。可以挑 provider,也可以给 system prompt。
  • ai_list — 看当前配了哪些路由、各自用哪个模型、有没有密钥(不吐密钥本身)。

它解决什么

DSH 自己的 Models 页管的是「你在输入框里选哪个模型跟你对话」。这个插件管的是另一件事:鲸鲸在干活的时候,能不能顺手把某个子问题甩给另一个模型——写文案、翻英文、做个第二意见、换个模型的视角挑错。

外部模型看不到当前会话,只能看到你传进去的那段 prompt。这不是缺陷,是设计:一次问答,账算得清,上下文不串味。

安装

# 在 DSH 的 profile 目录里
cd ~/.dsh/profiles/desktop
node pnpm.mjs add link:E:/path/to/dsh-ai-relay

然后确认 package.json 的 dsh.profile.bundles 里有 dsh-ai-relay,重启 DSH 生效。

配置

在 ~/.dsh/profiles/desktop/cordis.patch.yml 里加(或直接改本插件自带的 patch):

- insert:
    - id: ai-relay
      name: 'dsh-ai-relay'
      config:
        defaultProvider: gpt
        keysFile: C:/Users/you/.dsh/ai-relay.keys.json
        providers:
          - id: gpt
            label: GPT-4o mini
            api: openai
            baseURL: https://api.openai.com/v1
            model: gpt-4o-mini
            apiKeyEnv: OPENAI_API_KEY
          - id: claude
            label: Claude
            api: anthropic
            baseURL: https://api.anthropic.com
            model: claude-sonnet-4-5
            apiKeyEnv: ANTHROPIC_API_KEY
          - id: my-gateway
            label: 自建网关
            api: openai
            baseURL: https://gateway.example.com/v1
            model: some-model
            apiKeyEnv: GATEWAY_API_KEY
            headers:
              x-org: whale

字段说明:

字段必填含义
id是路由名,ai_ask 的 provider 参数就是它
api否openai(默认,聊天补全)或 anthropic
baseURL是写到 /v1 为止;插件自己拼 /chat/completions 或 /messages
model是请求里的模型 id
apiKeyEnv否从哪个环境变量取密钥
headers否额外请求头(网关鉴权、组织 id 之类)

密钥

两条路,都不进配置文件:

  1. 环境变量:路由的 apiKeyEnv 指向的那个变量。
  2. 本地密钥文件:默认 ~/.dsh/ai-relay.keys.json,内容按路由 id 索引:
{ "gpt": "sk-...", "claude": "sk-ant-..." }

找密钥的顺序:环境变量 → 密钥文件。都没有时 ai_ask 直接返回 ok:false 并说明该配哪个变量,不会发一个没鉴权的请求出去。

用起来长什么样

ai_ask(prompt: "把这段中文翻成自然的英文,保留技术词:……", provider: "claude")
→ [claude / claude-sonnet-4-5 · 1840ms · tokens 312→418]
   ...

失败时返回 ok:false 加一句人话提示(密钥被拒 / 端点 404 / 超时),不抛异常打断整轮。

参数

ai_ask:

  • prompt(必填,上限 20 万字符)
  • provider — 路由 id;不填用 defaultProvider,再不行用第一条
  • system — 可选系统提示
  • max_tokens — 上限 8192,Anthropic 默认 1024
  • temperature

ai_list:无参数。

测试

node test/test.mjs

自带的测试会起一个本地 mock 服务,模拟 OpenAI 与 Anthropic 两种响应,覆盖成功、缺密钥、未知 provider、401、坏配置。不联网、不需要真密钥。

边界

  • 只做「一问一答」,不转发工具调用,也不续写上下文。
  • 不替你保存历史;要留档就把回答写进你自己的笔记。
  • 超时默认 120 秒,timeoutMs 可调(5 秒到 10 分钟)。

MIT.