dsh-generate-image
Run the following command in DeepSeek Harness:
dsh plugin install zemul/dsh-generate-image
Paste the following prompt into your AI chat to install this plugin:
Install in DeepSeek Harness by running dsh plugin install zemul/dsh-generate-image; the source is at https://github.com/zemul/dsh-generate-image with zero dependencies and no build steps required.
About this plugin
Text-only models can often interpret images, yet they still lack a clean way to produce or edit one inside a harness workflow. dsh-generate-image adds that missing capability for DeepSeek Harness by exposing a generate_image tool that accepts a prompt for text to image creation or a prompt plus an input image for instruction driven editing, such as changing backgrounds, adjusting colors, removing local objects, or redrawing selected areas.
The plugin runs through the local Codex CLI image generation path, so it does not require an extra API key and does not bring npm dependencies or build steps. It locates newly generated files through directory snapshot diffs and respects timeout or abort signals, making it practical for automated harness sessions.
It is a good fit for developers, designers, and content creators who want image creation and editing inside an existing model workflow without standing up a separate image service.
Use Cases
- Generate product concept images directly in a Harness conversation from text prompts
- Pass in an existing design and instruct the model to swap backgrounds, adjust colors, or erase local elements
- Batch-call image generation in automated workflows without maintaining a separate image service
Best For
- Developers who want to embed image capability into their existing model workflow
- Independent designers and content creators needing quick prototyping
- Lightweight tool users who want to avoid extra API keys or npm dependencies
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.