DSH Hub
ModLens cover

liustack/modlens

ModLens

BundleVision632 GitHub stars

The first vision plugin for DeepSeek Harness. Pasted images become structured evidence: OCR, layout, entities. A text-only model can read a screenshot.

Install

npx -y @deepseek-ai/dsh plugin --profile web add @liustack/modlens

After install, pick the (modlens vision) variant in the model selector so pastes take the evidence path.

Related plugins