Agent skill

Paper Staged Gen

by Yrzsh in Yrzsh/paper-workbench

A skill your agent uses when generating or regenerating a full manuscript with the DSH Paper Workbench staged pipeline — contract-first, per-section generation with citation-pool gates, programmatic…

MITAuto-check passedResearch & Science

Install Paper Staged Gen

skills CLI
$ npx skills add Yrzsh/paper-workbench --skill paper-staged-gen -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install Yrzsh/paper-workbench paper-staged-gen --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/Yrzsh/paper-workbench.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/paper-staged-gen .claude/skills/paper-staged-gen && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
paper-staged-gen
GitHub stars
112
Token cost
~524 tokens
SKILL.md length
85 words
Files
1
Skills in repo
3
Repo updated
First seen
Licence
MIT

At a glance

A skill your agent uses when generating or regenerating a full manuscript with the DSH Paper Workbench staged pipeline — contract-first, per-section generation with citation-pool gates, programmatic…

  • Works in 5 steps: 契约先行,锁定才写:draft/contract.md… → 引文池硬约束:每段只能引用契约「引文分配」中分给该段的文献编号。段级门禁会自动拦截… → 表格只产数据不产排版:表格一律 generate… → …
  • Regenerating a full manuscript with the DSH Paper Workbench staged pipeline — contract-first
  • SKILL.md covers 何时触发, 不可违反的纪律, 标准编排序列(S0→S8) and 段级门禁三件套(自动,勿绕过), plus 2 more sections
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

Paper Staged Gen is an agent skill from Yrzsh/paper-workbench. Use when generating or regenerating a full manuscript with the DSH Paper Workbench staged pipeline — contract-first, per-section generation with citation-pool gates, programmatic tables, post-hoc abstract. Routes the wb.py generate subcommands; never one-shot a long manuscript.

Its SKILL.md is about 520 tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It sits in Research & Science, covering Citation management. The repository describes itself as: Paper Workbench — 论文写作审查全流程工作台 / End-to-end academic writing & review workbench. The licence is MIT.

When your agent uses it

  • Regenerating a full manuscript with the DSH Paper Workbench staged pipeline — contract-first
  • Per-section generation with citation-pool gates
  • Programmatic tables
  • Post-hoc abstract

Example prompts

  • “/paper-staged-gen”

Workflow steps

5 steps, taken from the first numbered list in SKILL.md.

  1. 契约先行,锁定才写:draft/contract.md 未"已锁定"前,禁止生成任何正文段落。锁定是人工检查点(S0),你只能提示用户去锁定,不能替用户锁定。
  2. 引文池硬约束:每段只能引用契约「引文分配」中分给该段的文献编号。段级门禁会自动拦截池外引用——被拦时不要硬塞,回头检查分配是否合理。
  3. 表格只产数据不产排版:表格一律 generate tables(JSON→管道表)。你只负责产出行数据(或 --extract 迁移存量),绝不手写表格排版。
  4. 摘要后置:摘要在全部正文段落定稿后用 generate abstract 生成,基于各段摘要而非全文;对齐校验失败就修,不要绕过。
  5. 失败重试有上限:单段门禁失败自动重试 1 次,仍失败则停下报告用户,不要无限循环。

What it can do on your machine

Read from SKILL.md and the folder at commit 801516a. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md.

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Paper Staged Gen loads about 524 tokens when it runs. Until then it costs about 74 tokens; SKILL.md has 85 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~74
When it runs · the whole SKILL.md, loaded when a task matches
~524

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from Yrzsh/paper-workbench at commit 801516a, republished under its MIT licence (© Yrzsh). 85 words, ~524 tokens.

Download SKILL.mdSave it as .claude/skills/paper-staged-gen/SKILL.md (or your agent's skills folder).
name
paper-staged-gen
description
Use when generating or regenerating a full manuscript with the DSH Paper Workbench staged pipeline — contract-first, per-section generation with citation-pool gates, programmatic tables, post-hoc abstract. Routes the `wb.py generate` subcommands; never one-shot a long manuscript.

Paper Staged Generation (paper-staged-gen)

分批生成编排协议。配套引擎:~/.dsh/papers/workbench/staged_gen.py + wb.py generate。 设计见工作区《分批生成方案》。核心信条:长文/多图表稿件绝不一次性整篇生成。

何时触发

  • 用户要"写一篇综述/长文"或"重写 main.md",且目标稿 > 3000 词或含 ≥2 张图表
  • 一次性生成的稿件出现表格损坏、图文分离、幻觉引用、摘要漂移,需要重写
  • 已有 draft/contract.md,要继续/恢复分段生成

不可违反的纪律

  1. 契约先行,锁定才写:draft/contract.md 未"已锁定"前,禁止生成任何正文段落。锁定是人工检查点(S0),你只能提示用户去锁定,不能替用户锁定。
  2. 引文池硬约束:每段只能引用契约「引文分配」中分给该段的文献编号。段级门禁会自动拦截池外引用——被拦时不要硬塞,回头检查分配是否合理。
  3. 表格只产数据不产排版:表格一律 generate tables(JSON→管道表)。你只负责产出行数据(或 --extract 迁移存量),绝不手写表格排版。
  4. 摘要后置:摘要在全部正文段落定稿后用 generate abstract 生成,基于各段摘要而非全文;对齐校验失败就修,不要绕过。
  5. 失败重试有上限:单段门禁失败自动重试 1 次,仍失败则停下报告用户,不要无限循环。

标准编排序列(S0→S8)

wb.py generate init                       # 初始化 draft/(仅首次)
wb.py generate contract [--ai]            # 生成契约 → 交用户补全并锁定(S0 人工检查点)
   ── 等待用户把「契约状态:待锁定」改为「已锁定」──
wb.py generate status                     # 查看段级进度,确定下一个待写段
wb.py generate section --sid S<N>         # 逐段生成(自动门禁+重试);重复直到全部 done
wb.py generate tables [--extract|--gen]   # 表格:迁移存量 / 渲染 / 委托产数据
wb.py generate abstract                   # 摘要后置生成 + 对齐校验
wb.py generate assemble                   # 拼装(自动插图插表、前置摘要)→ manuscript/main.md

每段生成后必须跑 generate status 确认 done 再进下一段;出现 failed 先看 issues 再决定重试或人工。

段级门禁三件套(自动,勿绕过)

  • 占位符/双标点/超长句(复用 toolbox.mechanical_check 的 P0/P1)
  • 引文范围 ⊆ 本节分配池(幻觉引用段级拦截)
  • 字数预算 ±20%

决策快捷

  • "这段引用了不在分配池的文献" → 查契约引文分配,缺则补分配并提醒用户重新锁定,不要删引用硬凑
  • "表格又乱了" → 一定是有人手写了表格;改走 generate tables JSON 渲染
  • "摘要说了正文没有的" → 对齐校验会拦;让摘要回退到各段摘要重写,不要从全文"发挥"
  • "能不能一次生成完" → 不能。除非稿件 < 2500 词且无图表,否则坚持分段

与其它技能的关系

  • 写作风格仍从 quality_rules.md + build_stage_prompt 取;本技能只管生成编排与门禁,不改写作风格源
  • 审查仍走 review-auto 门禁,本技能不替代审查
  • 期刊格式以 journal/chosen.md 为准,跨刊不得套用其它期刊预算

© Yrzsh, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in skills/paper-staged-gen of Yrzsh/paper-workbench.

Open the folder on GitHubat commit 801516a

Compare with similar skills

Paper Staged Gen next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Paper Staged Gen compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Paper Staged Gen this skillYrzsh/paper-workbench112—~524Automated safety check: PassMIT
Content Research Writerweapp-tailwindcss/weapp-tailwindcss1.9k25 repos~3.5kAutomated safety check: PassMIT
Systematic Review ScreenerImbad0202/academic-research-skills51k—~8.4kAutomated safety check: PassCustom licence
NetworkxzLanqing/codex-claude-academic-skills4.7k15 repos~3.2kAutomated safety check: PassBSD-3-Clause
Literature Reviewneflibata-feng/MyArxiv-Agent12620 repos~5.9kAutomated safety check: NotesMIT
Openalex Databaseneflibata-feng/MyArxiv-Agent12612 repos~3kAutomated safety check: PassCustom licence

Similar skills

  • Content Research Writer

    weapp-tailwindcss/weapp-tailwindcss

    Assists in writing high-quality content by conducting research, adding citations, improving hooks, iterating on outlines, and providing real-time feedback on each section.

    1.9k GitHub starsUsed in 25 repos~3.5k tokens
    Research & ScienceAuto-check passed
  • Systematic Review Screener

    Imbad0202/academic-research-skills

    Screens records for systematic, scoping and rapid reviews against fixed eligibility rules, using two blinded AI reviewers and a third adjudicator, with traceable PRISMA counts.

    51k GitHub stars~8.4k tokensUpdated 2 days ago
    Research & ScienceAuto-check passed
  • Networkx

    zLanqing/codex-claude-academic-skills

    Comprehensive toolkit for creating, analyzing, and visualizing complex networks and graphs in Python.

    4.7k GitHub starsUsed in 15 repos~3.2k tokens
    Research & ScienceAuto-check passed
  • Literature Review

    neflibata-feng/MyArxiv-Agent

    Conduct comprehensive, systematic literature reviews using multiple academic databases (PubMed, arXiv, bioRxiv, Semantic Scholar, etc.).

    126 GitHub starsUsed in 20 repos~5.9k tokens
    Research & ScienceAuto-check: notes
  • Openalex Database

    neflibata-feng/MyArxiv-Agent

    Query and analyze scholarly literature using the OpenAlex database.

    126 GitHub starsUsed in 12 repos~3k tokens
    Research & ScienceAuto-check passed
  • Citation Verification Guide

    Galaxy-Dawn/claude-scholar

    Reference guidance for checking every citation in academic writing against canonical sources such as DOI, arXiv, CrossRef and Semantic Scholar, to catch fake or wrong references.

    5.7k GitHub starsUsed in 2 repos~1.9k tokens
    Research & ScienceAuto-check passed

More from Yrzsh/paper-workbench

  • Scientific Writing

    Yrzsh/paper-workbench

    A skill your agent uses when writing or revising scientific manuscripts, abstracts, figures, or references for journal submission and you need full-paragraph prose, scientific structure…

    112 GitHub stars~7.4k tokensUpdated 1 mo ago
    Auto-check passed
  • Write Scientific Manuscript

    Yrzsh/paper-workbench

    Diagnose and repair the clarity and logic of manuscript prose, journal-agnostic.

    112 GitHub stars~2.7k tokensUpdated 1 mo ago
    Auto-check passed

Questions about Paper Staged Gen

What does Paper Staged Gen do?

A skill your agent uses when generating or regenerating a full manuscript with the DSH Paper Workbench staged pipeline — contract-first, per-section generation with citation-pool gates, programmatic…. Paper Staged Gen is an agent skill from Yrzsh/paper-workbench. Use when generating or regenerating a full manuscript with the DSH Paper Workbench staged pipeline — contract-first, per-section generation with citation-pool gates, programmatic tables, post-hoc abstract.

When should I use Paper Staged Gen?

Paper Staged Gen fits situations like: regenerating a full manuscript with the DSH Paper Workbench staged pipeline — contract-first; per-section generation with citation-pool gates; programmatic tables; post-hoc abstract.

How do I install Paper Staged Gen in Claude Code?

Run `npx skills add Yrzsh/paper-workbench --skill paper-staged-gen -a claude-code`. Or copy the skill folder (skills/paper-staged-gen in Yrzsh/paper-workbench) into .claude/skills/paper-staged-gen in your project. Claude Code loads it when a task matches its description.

How do I install Paper Staged Gen in Codex?

Run `npx skills add Yrzsh/paper-workbench --skill paper-staged-gen -a codex`. Or copy the skill folder (skills/paper-staged-gen in Yrzsh/paper-workbench) into .agents/skills/paper-staged-gen in your project. Codex loads it when a task matches its description.

Can I use Paper Staged Gen in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add Yrzsh/paper-workbench --skill paper-staged-gen -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/paper-staged-gen, .gemini/skills/paper-staged-gen, .github/skills/paper-staged-gen and .opencode/skills/paper-staged-gen in your project.

What does Paper Staged Gen need to run?

SKILL.md names no scripts, command-line tools or credentials: Paper Staged Gen is instructions for the agent only.

Does Paper Staged Gen access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Paper Staged Gen safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Paper Staged Gen use?

Paper Staged Gen is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Paper Staged Gen use?

About 524 tokens (SKILL.md is roughly 2.1k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Paper Staged Gen?

Skills that share tags, products or a category with Paper Staged Gen: Content Research Writer (weapp-tailwindcss/weapp-tailwindcss, 1.9k stars), Systematic Review Screener (Imbad0202/academic-research-skills, 51k stars), Networkx (zLanqing/codex-claude-academic-skills, 4.7k stars) and Literature Review (neflibata-feng/MyArxiv-Agent, 126 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Paper Staged Gen?

Yrzsh (a GitHub user) maintains it in Yrzsh/paper-workbench, which has 112 GitHub stars. The repository holds 3 skills in this directory. The repository was last updated on September 7, 2026.

Source: Yrzsh/paper-workbench on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.