visionpower
@piercehuang663
VisionPower: a portable image understanding MCP server for Codex, Claude Desktop, Cursor, and other agents.
dsh-vision-relay
About
Zero-modification, zero-switch DeepSeek Harness vision plugin: paste images for text-only models to read, automatic switching between cloud and local Ollama backends, ModLens v2-style structured evidence output.
Description machine translated from the original.
Topics
cordisdshimage-recognitionmultimodalocrollamatext-only-llmvision
Similar plugins
inkstone
@banlanzs
Browse and manage conversation history from Claude Code, Codex CLI, OpenCode, Reasonix, Oh My PI (omp), Pi, and DeepSeek Harness (dsh)
dsh-vision
@237229953-create
DSH plugin: text-only models (e.g. DeepSeek-V4) automatically see images via a vision model. Official surface-replace, cache-friendly, human transcript untouched. Text-only model automatic image recognition bridge