coder-wu/dsh-share
coder-wu★ 0JavaScript最后同步: 2026-08-16
—
README 摘要
dsh-share A DeepSeek Harness plugin that shares the current session page as a long screenshot — a single PNG of the whole conversation, not just the visible viewport. What it does Adds a copy button to the session header (top-right, next to the jobs / subagent buttons). Click it and the browser captures the full conversation — every loaded message — and copies it to the system clipboard as a PNG. Paste it anywhere: a chat, a document, an image editor. The page is painted by the browser's own layout engine into an SVG foreignObject (via modern-screenshot, bundled inline), so the result matches what you see: CSS variables, gradients, web fonts, and emoji render correctly. Dark themes keep their real background. Output is 2× device pixels (clamped to browser canvas limits for very long sessions). The capture is auto-cropped : the conversation column is capped ( max-width: 748px ) inside a wide container, so a full-container capture carries 200px of empty gutter on each side. A downscaled probe finds the content bounding box and the final image is trimmed to the content plus a small margin. The composer (input box) is excluded — it is full-width chrome that would otherwise keep the cro…
在 GitHub 查看完整 README →分类
The first vision plugin for DeepSeek Harness, and the vision bridge for every text-only coding agent. Paste an image, get structured JSON evidence (OCR, layout, semantics). | 全网最强 DeepSeek Harness 外挂视觉插件,为 DeepSeek、GLM 等纯文本模型外挂视觉能力,粘贴图片即得结构化 JSON 证据(OCR、版面、语义)。
★ 2,208
Anionex/agent-vision-toolkit为纯文本模型"看图“设计更好的视觉工具箱和技能,支持多图理解,图片问答,前端UI还原、GUI 自动化等,并可选无缝接入多个主流agent,直接识别粘贴图片| A vision toolkit and skill designed for text-only llms — image Q&A, long-screenshot OCR, frontend UI restoration, and GUI automation, with optional seamless integration for Codex, Claude Code, Pi, Oh My Pi, and OpenCode
★ 933
Anionex/dsh-vision-toolkit让纯文本模型更好地做视觉任务的DeepSeek Harness插件:带意图的图片问答、长截图 OCR、UI 还原等|DeepSeek Harness-native integration for agent-vision-toolkit: image Q&A, long-screenshot OCR, UI restoration, grounding, pixel diff, Artifacts, and Web UI.
★ 488