Bundle
@zenmux/dsh-plugins
Use ZenMux models in DeepSeek Harness with OAuth login, model search, and protocol-aware routing
- Source
- ZenMux
- stars
- 1 stars
- License
- MIT
- Updated
- Updated 5 days ago
Readme
# ZenMux for DeepSeek Harness [中文说明](https://github.com/ZenMux/dsh-plugins/blob/main/README.zh.md) · [GitHub](https://github.com/ZenMux/dsh-plugins) · [ZenMux](https://zenmux.ai) Use ZenMux models in DeepSeek Harness (DSH) without copying an API key. Sign in through your browser, search the full model catalog, and let the plugin choose the cache-capable protocol for each model. ## Features - Browser OAuth login with automatic token refresh. - 150 searchable text models in the current bundled catalog. - 126 models routed through Anthropic Messages with prompt-cache controls. - 24 OpenAI-only models routed through Chat Completions. - Every ZenMux inference request carries the current DSH session in `x-zenmux-session-id`. - `/zenmux login`, `/zenmux status`, `/zenmux logout`, and `/zenmux models refresh` commands. ## Install ```sh dsh plugin --profile web update @zenmux/dsh-plugins --latest dsh web ``` ## Sign in In a DSH conversation, run: ```text /zenmux login ``` Approve the ZenMux authorization page and return to DSH. Check the connection with `/zenmux status`; disconnect this installation with `/zenmux logout`. ## Choose a model Open the model selector and search by model name or ID. - **ZenMux · Anthropic**: cache-capable models using Anthropic Messages. Claude Fable 5 belongs here. - **ZenMux · OpenAI**: the remaining OpenAI-only models. The first eligible Anthropic request can create a prompt cache. Later requests with the same prefix can report cache reads. ## Update Quit the running DSH process, update the plugin, then restart DSH: ```sh dsh plugin --profile web add @zenmux/dsh-plugins@latest dsh web ``` Confirm the installed version with: ```sh dsh plugin --profile web list --depth 0 ``` Refresh the ZenMux model catalog (the command updates the settings file; restart DSH afterward): ```text /zenmux models refresh ``` The command fetches both ZenMux OpenAI and Anthropic listings and preserves OAuth credentials and other provider settings. ## Troubleshooting ### The latest plugin still shows only a few models An older manual model list is overriding the bundled catalog. In `~/.dsh/settings.yaml`, remove only the `models:` arrays under `llm-pi-ai.providers.zenmux` and `llm-pi-ai.providers.zenmux-models`, then restart DSH. Keep credential references and unrelated providers. ### `API key is invalid` or `AUTH` Run `/zenmux login` again. ### Network or TLS errors OAuth requests connect directly by default and can use an explicit HTTP(S)/SOCKS proxy. Custom TLS interception may also require `NODE_EXTRA_CA_CERTS` before DSH starts. See the [implementation and rollout notes](https://github.com/ZenMux/dsh-plugins/tree/main/docs/sdd) for advanced deployment details. ## Development ```sh pnpm install pnpm test ``` Model catalogs are release snapshots generated from ZenMux's OpenAI and Anthropic model-list endpoints. Source and issue tracking live at [ZenMux/dsh-plugins](https://github.com/ZenMux/dsh-plugins). MIT License
Install
dsh plugin --profile web add github:ZenMux/dsh-plugins
Profile: web
With the hub plugin installed, ask your agent to install it by name — it resolves the same plan shown here.
dsh plugin --profile web add github:stvlynn/dsh.fish#path:packages/dsh-plugin-hub
install zenmux-dsh-plugins from the hub
- This package builds from source on install. pnpm will ask you to allow its build script — that is permission to run the package’s code on your machine, outside the agent sandbox. Only allow sources you trust.
- This source has no pinned commit, so a later push upstream changes what installs. Prefer pinning a commit.