chou109/dsh-toolkit

chou109★ 0JavaScript最后同步: 2026-08-16

在 GitHub 打开

DeepSeek Harness 增强工具包,四个即插即用的小工具合而为一——粘贴图片自动识图、工作区一键在资源管理器/VS Code/终端打开并可复制路径、消息历史侧栏、已归档会话面板。一条命令安装全部组件。 a DeepSeek Harness enhancement pack — four tools in one: paste-image auto-vision, one-click workspace open (Explorer/VS Code/terminal) plus path copy, a message rail, and an archived-sessions panel. One command installs all.

README 摘要

dsh-toolkit A DeepSeek Harness enhancement pack — four plug-and-play tools in one : paste-image auto-vision, one-click workspace open plus path copy, a message-history rail, and an archived-sessions panel. One command installs everything. 国内镜像 / Mirror : also hosted on gitee.com/chill109/dsh-toolkit — Gitee is a mainland-China Git host (faster access from mainland China; use it if GitHub is slow). Components vision-bridge · workspace-launcher · msgrail · archives Platform Windows / macOS / Linux (dsh web) Requires DeepSeek Harness dsh web 0.1.0-rc.6, git , node License MIT 中文版见 README.zh.md。 Part 1 — If you are a human This part is written for people who just want everything installed. What the toolkit does One install gives you four enhancements: 1. vision-bridge — paste an image, get an answer. Paste any image into the chat box and send it, even with a text-only model: the harness hands the agent the image as a local path and the agent automatically calls its available vision tool ( vision chat from qwen-mm-plugins-api, or any other vision tool you configured) to describe/OCR/answer. No more "current model does not support images". 2. workspace-launcher — open the workspace, or c…

在 GitHub 查看完整 README →
终端/TUIdeepseekdeepseek-harnessdshdsh-plugindsh-pluginsterminalvision

分类