Research Paper

What it does

When your agent uses it

• Tasks that involve Scientific writing
• Tasks that involve Literature review
• Tasks that involve Brainstorming

Example prompts

• “/research-paper”

Requirements

• Python 3
• A Bash shell

Workflow steps

8 steps, taken from the step headings in SKILL.md.

1. 1Idea Agent（选题）
2. 2Literature Agent（文献）
3. 3Method Agent（方法设计）
4. 4Experiment Agent（实验）
5. 5Writing Agent（写论文）
6. 65: Figures Agent（论文绘图）
7. 7Review Agent（模拟审稿）
8. 8Rebuttal Agent（反驳）

What it can do on your machine

Read from SKILL.md and the folder at commit 583700e. 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

  Ships 1 file in scripts/ (Shell, from the files we listed), which the agent can run.

  Shell commands in SKILL.md call:

  • python
  • curl
  • python3
  • bash

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

• Network

  Hosts in commands or code, which the agent is likely to contact:

  • export.arxiv.org
  • api.semanticscholar.org

  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

Research Paper loads about 1.9k tokens when it runs, and up to ~18k if it reads all its reference files. Until then it costs about 96 tokens; SKILL.md has 503 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it

~96

When it runs · the whole SKILL.md, loaded when a task matches

~1.9k

With references · SKILL.md plus every file in references/, read only if the agent opens them

~18k

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

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); the scripts in this folder are not scanned.

SKILL.md

The full file from SGloria/research-paper-pipeline at commit 583700e, republished under its MIT licence (© SGloria). 503 words, ~1,918 tokens.

name

research-paper

description

CS 顶会论文写作全流程 Skill（7-Agent Pipeline）。覆盖从选题到投稿的完整管线：Idea 生成、Literature Review、Method 设计、Experiment 方案、Paper Writing、模拟审稿、Rebuttal。 触发词包括但不限于：写论文、发顶会、research paper、论文选题、idea generation、literature review、模拟审稿、rebuttal、paper writing、帮我写 paper、投稿、选题、novelty check。 当用户想要系统性地撰写一篇学术论文，或者需要某个阶段的帮助（如只做选题、只做文献综述、只做模拟审稿）时触发。 不要用于公众号写作（用 khazix-writer）、不要用于产品调研（用 hv-analysis）。

mkdir -p "[项目名称]"/{sections,figures/svg_output,figures/output,data,output}

[项目名称]/
  research_state.json      # 全局状态文件（断点续跑）
  idea_output.json         # Idea Agent 输出
  literature_pool.json     # 文献池
  references.bib           # BibTeX 文件
  sections/                # 各 section 的 Markdown 草稿
    abstract.md
    introduction.md
    related_work.md
    method.md
    experiments.md
    conclusion.md
  figures/                 # 图表文件
    figure_list.json       # 图表清单（id/type/description/status）
    svg_output/            # SVG 源文件
    output/                # 导出产物
      paper_figures.pptx   # 所有 SVG 汇总的 PPTX
  data/                    # 实验数据
  output/                  # 最终产出
    paper.md               # 完整 Markdown 论文
    paper.tex              # LaTeX 版本
    paper.pdf              # 编译后 PDF
    rebuttal.md            # Rebuttal letter

{
  "project_name": "",
  "target_venue": "",
  "current_phase": "idea",
  "selected_idea_index": null,
  "phases": {
    "idea": {"status": "pending", "output_path": null},
    "literature": {"status": "pending", "output_path": null},
    "method": {"status": "pending", "output_path": null},
    "experiment": {"status": "pending", "output_path": null},
    "writing": {"status": "pending", "output_path": null},
    "review": {"status": "pending", "output_path": null},
    "rebuttal": {"status": "pending", "output_path": null}
  },
  "created_at": "",
  "updated_at": ""
}

python3 [skill目录]/scripts/figure_to_pptx.py "[项目路径]/figures"

python [skill目录]/scripts/md_to_latex.py \
  "output/paper.md" \
  "output/paper.tex" \
  --venue <venue_name> \
  --bib "references.bib"

bash [skill目录]/scripts/build_paper.sh "output/paper.tex"

python [skill目录]/scripts/verify_citations.py "references.bib"

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

Files

SKILL.md and 32 other files (scripts, references, assets) in the repository root of SGloria/research-paper-pipeline.

• SKILL.md
• .gitignore
• LICENSE
• README.md
• assets/wechat-pay.jpg
• references/citation-verification.md
• references/experiment-agent.md
• references/figures-agent.md
• references/host-integration.md
• references/idea-agent.md
• references/literature-agent.md
• references/method-agent.md
• references/rebuttal-agent.md
• references/review-agent.md
• references/venue-templates.md
• references/writing-agent.md
• requirements.txt
• scripts/build_paper.sh
• … and 15 more

Open the folder on GitHubat commit 583700e