agent-vision-toolkit
Vision toolkit for text-only LLMs. Works with Codex, Claude Code, Pi, OpenCode — and tags dsh-plugin.
Install
Forwards to pnpm inside the web profile. Confirm the README if the author documents a .dsh-plugin path.
dsh plugin --profile web add github:Anionex/agent-vision-toolkit
What it is
The parent toolkit behind dsh-vision-toolkit. It is not Harness-only: the same image Q&A / OCR / GUI automation stack plugs into several agents. We list it because many dsh users install the generic toolkit first, then the Harness-native wrapper. If you only run DeepSeek Harness, prefer dsh-vision-toolkit. If you hop between agents, keep this one.
Source: Anionex/agent-vision-toolkit. Official harness: deepseek-ai/deepseek-harness.
Cross-agent toolkit; confirm the Harness adapter before relying on UI cards.
Related in Vision
First vision plugin for DeepSeek Harness. Paste an image, get structured JSON: OCR, layout, semantics.
Harness-native vision toolkit: image Q&A, long-screenshot OCR, UI restoration, grounding, pixel diff, Artifacts.