DeepSeek Harness plugin

paper-workbench

DSH bundle:注册 paper 工具(论文解析/工作台状态),工作流知识见 skills/paper-lab(由 /paper-init 装入工作区 .agents/skills/)

Jump to install

Source facts

Repository
LessXi/paper-workbench
Latest update
Aug 19, 2026
Category
Workflow & Automation
GitHub stars
0
Format
plugin
Catalog evidence
Upstream dsh.bundle evidence
Evidence path
package.json#dsh.bundle
Checked against
0.1.0-rc.8
Upstream check date
2026-08-20

This evidence comes from the upstream catalog. This site has not installed, run, or security-reviewed the plugin.

Install

Start with a prompt that asks an agent to review the GitHub repository and source. Switch to the command if you want to install it yourself.

Copy this prompt into DSH, Codex, or another agent and ask it to review the GitHub repository and source first.

Do not install or run any commands yet. Read this plugin's GitHub repository, README, and relevant source code. Then answer the questions below clearly and directly so I can decide whether it fits my needs:

1. What is this plugin, and what problem does it solve?
2. Who is it for, and what are its typical use cases?
3. How is it used after installation? Include one minimal example.
4. What known limitations or privacy, security, compatibility, or maintenance risks does it have?
5. Give a clear recommendation: recommend, conditionally recommend, or do not recommend, with reasons.

Distinguish statements documented by the repository, inferences from source code, and unknowns. If evidence is insufficient, say so explicitly. Do not guess or simply repeat the README.

GitHub: https://github.com/LessXi/paper-workbench
Plugin: paper-workbench
Author: LessXi

Check the source files

Read the README and other files from this plugin directory before installing.

File explorer3 files
README.mdSource · read only

paper-workbench

把你的 agent(ZCode / DeepSeek Harness / Claude Code)变成全流程论文研读学习复现工作台

发现 ─ 获取 ─ 解析 ─ 精读 ─ 沉淀 ─ 复现 ─ 综述
find   read   parse  learn  vault  repro  survey

特色是学习模式(/paper-learn):AI 不代读——只做导读地图、答疑、复述纠错、闭卷出题、code review,公式和代码过你自己的脑子。

安装

方式一:插件市场(推荐,零终端)——ZCode/Claude Code:设置 → 插件管理 → Discover → + 从 GitHub 添加 https://github.com/LessXi/paper-workbench → 安装。装完重开会话。

方式二:DSH 一条命令(bundle)——DeepSeek Harness 用户:

dsh plugin --profile web add github:LessXi/paper-workbench   # 在 DSH 源码目录跑 pnpm dsh …,可加 #<commit> 锁定

装完获得 paper 工具(status 查工作台状态 / parse 解析 PDF);工作流技能由 /paper-init 自动落到工作区 .agents/skills/

方式三:一条 bootstrap 命令(其他 terminal 用户)

curl -LsSf https://raw.githubusercontent.com/LessXi/paper-workbench/main/bootstrap.sh | bash
# Windows: powershell -ExecutionPolicy Bypass -c "irm https://raw.githubusercontent.com/LessXi/paper-workbench/main/bootstrap.ps1 | iex"

(备选:git clone 本仓库后跑 install.sh / install.ps1

两步跑通

1. 插件装好,重开会话,在任意目录说 /paper-init   ← 生成你的工作台(只问一个问题:解析环境 lite 还是 full)
2. /paper-find <你的主题>                        ← 检索落库,开始研读

解析环境两档:lite 秒级装完(纯文本抽取,够筛选);full 公式还原 LaTeX、表格成表、抽全部图表(精读必备),首次多等十来分钟。lite 随时可升级 full(一条 pip 命令,parse_paper.py 会在需要时提示你)。

九个命令

命令用途
/paper-init [目录] [--lite]初始化工作台(--lite 秒级轻量环境)
/paper-find <query>多源检索 → 必读/备读/跳过分档 → 落库登记
/paper-read <论文>AI 代读:MinerU 解析 + 三遍读法 + 结构化笔记
/paper-learn <论文>学习模式:你读你写,AI 陪读/出题/review
/paper-cite <论文>前向/后向引文图谱 + BibTeX
/paper-survey <topic>方法族谱 + 对比矩阵 + 演化时间线 + canvas
/paper-watch <topic>arXiv 订阅追新 digest
/paper-reproduce <论文>CPU 级复现:官方repo→降规模→从零实现→数值对照
/paper-lab台账盘点、卡点诊断、下一步建议

依赖

  • 必需uv(Python 3.12 自动管理)
  • 增强:arxiv-mcp-server + paper-search-mcp(/paper-init 自动写入工作区 MCP 配置;无 MCP 环境自动降级为网页检索 + curl 直拉);Obsidian(工作区本身就是 vault——明面全是论文数据,Obsidian Sync 直接全同步;框架收在 .zcode/ 点目录,Obsidian 不碰,无需手动排除)
  • 解析:MinerU(公式/表格/中文最强,CPU 可跑)+ pypdf 快速兜底

多机环境坑(杀软拦 uv trampoline、uvx 默认 Python、mineru 漏 six 等)已沉淀在 [skills/paper-lab/references/environment-notes.md](skills/paper-lab/references/environment-notes.md),安装脚本内置规避。

结构

skills/paper-lab/            自包含技能包(SKILL.md + 9 命令 + 模板 + 脚本 + 踩坑档案)
.claude-plugin/              插件清单(ZCode / Claude Code 市场兼容)
install.sh / install.ps1     一键双环境安装

License

MIT