Skip to content
  • Home
  • Plugin manual
  • dsh-commandcode-go-provider

jiesou/dsh-commandcode-go-provider

30Last commit Aug 15, 2026

dsh-commandcode-go-provider DSH plugin

This plugin bridges the Command Code Go/GOAT/Pro Plan subscription to DSH's native LlmAdapter. It dynamically fetches the model list from /provider/v1/models, filters according to Go Plan rules, and merges Reasoning Effort support from the official CLI catalog. After installation, you can select the Command Code Go provider and its models in the DSH Web interface.

developer-tools

How to install the dsh-commandcode-go-provider DSH plugin

dsh plugin --profile web add @jiesou/dsh-commandcode-go-provider

Copying does not run this command. Review the repository and version before installing the dsh-commandcode-go-provider DSH plugin.

dsh-commandcode-go-provider DSH plugin data source

dsh-commandcode-go-provider DSH plugin snapshot date: Aug 16, 2026

discovered

What the dsh-commandcode-go-provider DSH plugin can do

  • Streams LLM responses via the Command Code `/alpha/generate` endpoint
  • Auto-fetches and filters model list based on Go Plan rules (keeps open-source models, includes a few premium exceptions)
  • Merges Reasoning Effort support per model from the official CLI catalog (CDN)
  • Supports configurable maxTokens (default 64000) and defaultContextWindow (default 1,000,000)
  • Works with DSH's native LlmAdapter without additional configuration beyond the API key

Where the dsh-commandcode-go-provider DSH plugin fits

  • Use Command Code Go/GOAT/Pro Plan subscriptions directly in DSH without needing the Command Code CLI
  • Access open-source models like DeepSeek, Qwen, MiniMaxAI, etc. via DSH's model selection interface
  • Leverage premium models included in the Go Plan (e.g., GPT-5.6 Luna, Grok 4.5, Muse Spark 1.2 Contributor)
  • Automatically get updated model lists without manual configuration
  • Integrate reasoning effort control (off, minimal, low, medium, high, max) when supported by the model

Who the dsh-commandcode-go-provider DSH plugin is for

  • Command Code Go/GOAT/Pro Plan subscribers who want to use those models inside DSH
  • DSH users who prefer a plugin-based LLM provider over direct OpenAI-compatible endpoints

dsh-commandcode-go-provider DSH plugin limitations

  • Only works with Command Code's Go/GOAT/Pro Plan subscriptions; the general Provider API is not supported (returns 403 upgrade_required)
  • Requires a valid Command Code API key (user_xxxx) configured in `~/.dsh/.credentials.yaml`
  • Premium models from providers like Claude, Gemini are excluded from the model list
  • Depends on the availability of Command Code's /alpha/generate and /provider/v1/models endpoints

dsh-commandcode-go-provider DSH plugin: from the repository README

Quoted from the jiesou/dsh-commandcode-go-provider README, the upstream source of the dsh-commandcode-go-provider DSH plugin. Copyright remains with the original authors.

Command Code Go API provider for dsh. [English](README.en.md) Command Code 提供的订阅分两种： 1. **Provider API**：提供标准 OpenAI 兼容端点，可以直接接入任何 agent harness，不需要第三方插件。 2. **Go / GOAT / Pro Plan**：调用 Provider API 端点会返回 `403 upgrade_required`，只能通过 Command Code 私有的 CLI 网关 `/alpha/generate` 使用（vendor lock-in）。 本插件针对第二种情况：通过 `/alpha/generate` 流式接入 DSH 的原生 `LlmAdapter`，让 Go / GOAT / Pro Plan 用户直接在 DSH 中使用订阅的模型。模型列表不写死在代码里，插件会定时从 `/provider/v1/models` 拉取实时目录，并按 Go Plan 的规则筛选： - **默认保留开源模型**（deepseek、Qwen、MiniMaxAI、xiaomi、stepfun、tencent、nvidia、moonshotai 等 provider）。 - **包含几个 Go Plan 的 premium model 例外**，如 GPT-5.6 Luna、Grok 4.5、Muse Spark 1.2 Contributor。 - 其余 premium 模型（Claude、Gemini 等）一律排除。 筛选后还会从官方 CLI catalog (CDN) 合并每个模型的 Reasoning Effort 支持。 ## 安装 从 npm 安装（预构建产物，推荐）： ```sh dsh plugin --profile web add @jiesou/dsh-commandcode-go-provider ``` 或从 GitHub 安装： ```sh dsh plugin --profile web add github:jiesou/dsh-commandcode-go-provider ``` ## 安装之后 Command Code 的 API Key 应写入 `~/.dsh/.credentials.yaml`： ```sh echo 'COMMANDCODE_API_KEY: [your key, be like user_xxxx]' >> ~/.dsh/.credentials.yaml ``` 模型列表 **无需任何配置** ，插件启动后会自动从 `/provider/v1/models` 同步你的 Go 计划包含的模型，并从官方 CLI catalog (CDN) 合并每个

Read the full READMERepository license: MIT

dsh-commandcode-go-provider DSH plugin questions

How do I install this plugin?

Run `dsh plugin --profile web add @jiesou/dsh-commandcode-go-provider` from the command line. This will install the prebuilt npm package. Alternatively, you can install from GitHub with `dsh plugin --profile web add github:jiesou/dsh-commandcode-go-provider`.

Where do I put my Command Code API key?

Add it to `~/.dsh/.credentials.yaml` in the format `COMMANDCODE_API_KEY: user_xxxx`. Replace `user_xxxx` with your actual key. The plugin reads the key from the environment variable name specified in the config (default `COMMANDCODE_API_KEY`).

Why don't I see all models in the list?

The plugin filters models according to the Go Plan rules: it keeps open-source models (e.g., DeepSeek, Qwen, MiniMaxAI) and a few premium exceptions (e.g., GPT-5.6 Luna, Grok 4.5). Models from providers like Claude, Gemini are excluded. If you have a different plan, you may need to use the Provider API or a different plugin.

How do I update the model list?

The plugin automatically fetches the model list from `/provider/v1/models` on startup and periodically. You don't need to manually update. If you want to force a refresh, restart the DSH plugin or the DSH agent.

Can I adjust the max tokens or context window?

Yes, you can configure `maxTokens` (default 64000) and `defaultContextWindow` (default 1,000,000) in the plugin config. All other parameters are optional. Just add them to your DSH profile under the plugin's config section.