dsh-commandcode-goat: Integrate Command Code Subscriptions as an llm-pi-ai Route

Introduction

DeepSeek Harness (DSH) supports extending model providers through plugins. Command Code offers four subscription tiers: Go / GOAT / Pro / Max, which usually require separate management of API keys and usage. This plugin publishes these subscriptions uniformly to DSH’s llm-pi-ai namespace and provides account usage display and web search functionality. All operations use the same credentials.

Plugin Information

  • Name: dsh-commandcode-goat
  • Author: Plocr
  • Category: Model Inference
  • License: MIT

Core Features

  • Publishes Command Code subscriptions (Go / GOAT / Pro / Max) as llm-pi-ai provider routes.
  • Displays account usage (5-hour / weekly / monthly windows).
  • Enables web search using the same credentials.
  • Automatically syncs providers and cleans up unused routes.
  • Supports management of multiple subscription tiers.

Installation

Install the plugin using the following command:

dsh plugin --profile web add link:<本仓库根目录>

Restart the profile after installation.

Usage Steps

  1. Wait for automatic sync: After restarting the profile, wait a few seconds. commandcode-goat-autosync will automatically appear in the Settings → Models list.
  2. Enter the API key: In the generated provider entry (e.g., Command | goat), enter your API Key.
  3. Manage configuration: Use the provider tab to manage subscription tiers, toggle web search, and view usage.

Technical Details

  • Adapter implementation: This plugin does not implement LlmAdapter; instead, it writes configuration directly to DSH’s llm-pi-ai namespace, which is served by DSH’s built-in adapter.
  • Version requirement: Requires DSH version >= 0.1.7-alpha.2.
  • Sync mechanism: The sync operation is idempotent and preserves user-configured apiKeyEnv, baseURL, and modelOverrides; it does not overwrite values changed by the user.

Conclusion

This plugin aims to simplify management of Command Code’s multi-tier subscriptions by seamlessly integrating them into DSH’s model routing system. For more details, refer to the dsh.so catalog or the GitHub repository.

羽毛球分组比赛记分
小程序二维码

欢迎使用《羽毛球分组比赛记分》微信小程序

Xiaoye