Back to directory

dsh-image-gen

A DeepSeek Harness plugin that adds a full AI image workflow to the chat interface: text-to-image, continuous editing, a drawing canvas, batch generation, model comparison and a local ComfyUI connection.

dsh-image-gen

dsh-image-gen extends DeepSeek Harness (DSH) with conversational image generation and iterative editing, an AI creation canvas for drafting and placing reference images, a Studio workspace for batch creation with multiple reference images and advanced parameters, side-by-side comparison across models, a library of more than 500 prompts, and gallery management.

It supports cloud image models including Gemini, OpenAI-compatible endpoints, Seedream, DashScope, Grok Imagine and GLM-Image, as well as a local ComfyUI workflow imported as an API-format workflow JSON. A provider can be used with a person's own API key, or through an existing ChatGPT, Grok or Google subscription without a separate key, and generated results are stored isolated by workspace.

It installs into a DSH project with pnpm dsh plugin --profile web add dsh-image-gen@latest, and requires DSH's stable release with Node.js 22.19 or later, or 24 or later.

More in Agent skills

See all 581
  • Provenance gate

    github.com

    A deterministic authorization gateway that sits between an AI agent and its tools, tracing where each parameter of a side-effect call came from, and blocking the call or asking a human when the source does not check out.

  • A markdown framework of system rules that restates Isaac Asimov's Four Laws of Robotics as a set of prioritized loss calculations for large-language-model agents to follow, published as a discovery file for agents and crawlers to read.

  • Vitruvius

    github.com

    An engineering research agent that runs a discover, read, synthesize, verify and review loop over technical questions, reading sources directly rather than inferring from titles, and publishing its own scored verifier benchmark, weaknesses included.