DSH Plugins
返回列表
🧰

dsh-cherry-provider-bridge

联网工具 更新于 2026.09.08

在终端中运行以下命令:

dsh plugin install LiweiDonVee/dsh-cherry-provider-bridge

将以下提示词粘贴到 DeepSeek Harness 对话框中:

在 DeepSeek Harness 中执行 dsh plugin install LiweiDonVee/dsh-cherry-provider-bridge 完成安装,插件开源地址为 https://github.com/LiweiDonVee/dsh-cherry-provider-bridge

插件介绍

You have already tuned a provider in Cherry Studio—key, endpoint, model catalog—and you want to reuse exactly that setup in DeepSeek Harness without re-typing anything. dsh-cherry-provider-bridge does one job well: it reads Cherry Studio's SQLite database in read-only mode, pulls the credential and model route for the provider you choose, and writes them into DSH's credential store and route table. No GET /models discovery call, no database ever leaves your machine.

Synchronization ticks every five seconds, skips unchanged values, and retries on transient lock or namespace conflicts. It appends a missing /v1 suffix once, preserves model metadata and compatible reasoning options, and touches only the single provider you name—everything else stays untouched. Data sovereignty is yours: removing the plugin stops the sync loop but leaves the already-copied key and route in DSH; clean those up manually if you want.

This is for the developer who manages multiple providers through Cherry Studio and calls them through DeepSeek Harness, and would rather not paste API keys and model lists between two toolsets. It is an independent community adapter, not an official Cherry Studio plugin and not affiliated with DeepSeek AI, released under the MIT license.

使用场景

  • 在 DeepSeek Harness 中直接复用 Cherry Studio 里已配置好的 Provider 与模型目录,无需手动复制密钥和端点
  • 绕过 GET /models 接口发现机制,仅凭本地 SQLite 数据即可完成路由注册,降低对外部 API 的依赖
  • 多 Provider 环境下快速验证或切换模型,保留原始元数据与推理选项,避免二次配置丢失字段

适合人员

  • 同时使用 Cherry Studio 管理和 DeepSeek Harness 调用的开发者
  • 希望避免在两个工具间手动粘贴 API Key 与模型列表的运维或研究人员
  • 重视本地数据主权、不希望凭据经外部服务中转的隐私敏感用户