updated: dsh-vision-proxy

DeepSeek Harness released dsh-vision-proxy 0.1.3, a plugin that lets DeepSeek answer image-based questions by automatically translating attached images into text via Qwen VLM. The plugin, authored by Flyvhidbwo and licensed under MIT, is installable via 'dsh plugin --profile web add dsh-vision-proxy@0.1.3' and includes a CN mirror for offline installation.

dsh-vision-proxy DeepSeek Harness 插件：DeepSeek 大脑 + 自动识图。附加图片自动经 Qwen VLM 转译成文字后交给 DeepSeek 作答 dsh plugin --profile web add dsh-vision-proxy@0.1.3 Install with mirror acceleration xlings install dsh:dsh-vision-proxy -y Mirrored to xlings-res with a sha256 and a CN mirror: installs offline. Plugin - bundle dsh-vision-proxy - upstream Flyvhidbwo/dsh-vision-proxy - README profile web - license - MIT - bundle patch ./cordis.patch.yml Pinned commits - 0.1.0 bede819d322c - 0.1.3 1ce3d215c73d Credits - authors - Flyvhidbwo Keywords dsh ocr vision vlm