dsh-plugin-subs
Run the following command in DeepSeek Harness:
dsh plugin install randomix777/dsh-plugin-subs
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install randomix777/dsh-plugin-subs in your terminal to install this plugin; the full source is available at https://github.com/randomix777/dsh-plugin-subs .
About this plugin
You probably already pay for Claude Pro, ChatGPT Plus, X Premium, or a Qwen Coding Plan, yet the LLM power inside those subscriptions is locked behind each vendor's own web app or desktop client. DeepSeek Harness, where you do your real work, has no way to reach into those subscriptions unless you spin up separate API keys for each one. dsh-plugin-subs closes that gap by turning your existing consumer subscriptions into first-class LLM providers inside Harness, eliminating the API-key-and-env-var treadmill entirely.
The plugin ships OAuth or device-flow sign-in for nine providers: Claude (Pro/Max), Codex (ChatGPT Plus/Pro), Grok (X Premium), Google Antigravity (Google One AI Premium), OpenRouter, Agnes AI, Qwen Code, iFlytek Spark, and Baidu ERNIE. Most require zero environment variables: you complete a browser-based OAuth handoff and you are ready to chat. Once connected, you get streaming, tool calling, and image input wherever the upstream supports them. Active Grok sessions unlock three built-in tools—x_search, image_generate, and video_generate—while Codex adds its own image-generation path. Per-provider model catalogs are embedded and can be overridden in config so your setup stays predictable.
It is a natural fit if you do not want to maintain a second set of API keys for models you already subscribe to, if you want to A/B compare Claude, GPT, and Gemini side by side inside Harness, or if you simply want to reuse the login you already have in Claude Code. One OAuth tap, pick a model, and get back to building.
Use Cases
- Call Claude, Codex, Grok, and other subscription models from one harness
- Reuse your existing Claude Code login without extra configuration
- Side-by-side compare LLM outputs across different consumer subscriptions
Best For
- Developers holding multiple consumer AI subscriptions
- Harness users who want to minimize API key management overhead
- Individual developers needing a unified entry point for multi-model calls
Related Plugins
A unified suite combining hot runtime injection, task-aware thinking-mode routing, and a graded session protocol with red-team gates to sustain model diligence across long-horizon inference.
ModLens is a vision plugin for DeepSeek Harness that gives text-only models sight by reading images pasted directly into chat, with zero-config setup and multiple vision engines.
On-demand vision for text-only DeepSeek Harness agents: built-in free keyless vision chain and 14 vision tools, routing image turns as tool calls to vision models with pixel fidelity, no Python needed, one-command install.
Give text-only models in DeepSeek Harness eyes, enabling image Q&A, long-screenshot OCR, UI restoration, and GUI visual tasks.