dsh-codex
Run the following command in DeepSeek Harness:
dsh plugin install ddll8023/dsh-codex
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install ddll8023/dsh-codex in DeepSeek Harness to install the plugin; source is available at https://github.com/ddll8023/dsh-codex. Restart dsh web after installation to activate it.
About this plugin
dsh-codex solves a very practical gap: your DeepSeek Harness is already running, you also have a ChatGPT Plus or Pro subscription, but you do not want to request a separate DeepSeek API key just to use Codex models, nor juggle two credential systems. Install the plugin, complete one OAuth login, and the model picker instantly surfaces the full OpenAI Codex lineup with streaming text, tool calls, and image input all working out of the box, using the exact same event format as your existing providers.
Under the hood the plugin speaks the Codex Responses API rather than the standard /v1/chat/completions endpoint, unlocking a set of native capabilities: a per-session Fast mode that requests priority processing via the service_tier field for lower latency; native web search that runs inside the same response stream just like the official Codex CLI, with no extra round-trip and no dependence on DeepSeek search backends; and real-time quota readouts covering the 5-hour window, weekly window, and Spark sub-windows with reset countdowns and usage progress bars. The settings panel adds a dedicated Codex account section where browser login, device-code login, and usage refresh are all visual; the OAuth token never leaves the server-side credential store (0600 file permissions), appearing in no logs, no telemetry, and no session events.
Who is it for? If you already write code or handle everyday inference with DSH and hold a ChatGPT Plus or Pro subscription, and you want to switch between DeepSeek and Codex models within the same workflow; or if you care most about native search quality, fast response times, and quota transparency without configuring an extra Search Provider for Codex. One OAuth login covers everything, no additional API key required.
Use Cases
- Switch between DeepSeek and Codex models within the same DSH workflow without a separate API key
- Leverage native Codex web search for time-sensitive or factual queries without configuring third-party search backends
- Toggle Fast mode per session for priority processing while monitoring 5-hour and weekly quota via the usage panel
Best For
- Developers with a ChatGPT Plus/Pro subscription who already use DSH for daily inference
- Users who value search quality and low latency without setting up a separate Search Provider for Codex
- Ops teams who prefer one OAuth login to manage multi-model credentials instead of scattered API keys
Related Plugins
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.
An armor-breaking plugin for DeepSeek: rewrites instruction priority, expands refusal-blocked phrases, persists across sessions, and shows a green active indicator.