Long tasks shouldn't cost like they need a frontier model. Cheap-model long-task engine: reliability in structure (disk state machine + mechanical gates + T3 zero-prose dispatch + multi-brain audit), not in model strength. Pure Python stdlib. DSH skill.
Use your ChatGPT subscription in DeepSeek Harness to access Codex models and generate or edit images with gpt-image-2—no OpenAI Platform API key required.
Paste an image into the chat box and text-only DSH models can "see" it — auto-rewrite of pasted images + analyze_image tool routed to a Kimi vision model.
The auto permission tier for DeepSeek Harness: workspace-write requests run directly, sandbox escalations (danger-full-access) are judged by a separate classifier model.
Connect your sub2api gateway to DeepSeek Harness: OpenAI-compatible multi-provider routes (OpenAI / Claude / Grok / Gemini) behind one base URL, with per-key model discovery, usage lookup, and a settings page.
dsh plugin: model catalog auto-discovery - fetch model listings, pricing and capabilities from OpenAI-compatible API hosts, normalized into ready-to-use config