AgentSkillsCN

paper-tex-workflow

适用于编辑根目录下的 paper.tex 和 references.bib 文件（英文论文）。当您在仓库根目录下处理 paper.tex 文件、检查引用，或从 PDF 中提取论文内容时，可使用此技能。

SKILL.md
--- frontmatter
name: paper-tex-workflow
description: Workflow for editing the root-level paper.tex and references.bib (English papers). Use when working with paper.tex at repo root, checking citations, or extracting content from PDFs in papers/.
license: Complete terms in LICENSE.txt

paper.tex Workflow (Root English Paper)

Context

This repo has two distinct workflows:

  1. •TCC/ABNT review (PT-BR): work in project/ directory (see latex-tcc-abnt-review-ptbr skill)
  2. •English paper drafting: work with paper.tex and references.bib at repo root (this skill)

When to use this skill

  • •User asks to edit/draft/improve paper.tex or root-level English papers
  • •User requests citation checking or bibliography management for the root paper
  • •User wants to extract content from PDFs in papers/ directory

File locations

  • •Main paper: /paper.tex (repo root)
  • •References: /references.bib (repo root)
  • •PDF sources: /papers/*.pdf (optional)
  • •Citation checker: /tools/check_cites.py
  • •PDF extractors: /tools/extract_papers.py, /tools/review_papers.py

Core workflow steps

1. Citation Hygiene Check

Always run before major edits to understand citation state:

bash
python tools/check_cites.py

Output shows:

  • •Citations in paper.tex: count
  • •Entries in references.bib: count
  • •Missing bib entries: keys cited but not defined (fix these!)
  • •Unused bib entries: keys defined but never cited (optional cleanup)

2. Extract Content from PDFs (if applicable)

If papers/ contains PDFs of related work:

Quick abstract extraction:

bash
python tools/extract_papers.py

Shows title + abstract for each PDF.

Detailed section extraction:

bash
python tools/review_papers.py

Attempts to extract: title, abstract, dataset, method, results.

Use extracted content to:

  • •Draft Related Work section
  • •Add proper citations to references.bib
  • •Compare methods/results in Discussion

3. Edit paper.tex

Follow IMRAD structure (see latex-paper-writing-en skill for detailed guidance).

Key repo conventions:

  • •Use LaTeX quotes: ``like this'' not "like this"
  • •Foreign terms in italics: \textit{term}
  • •Citation format depends on loaded package (natbib/biblatex)

4. Manage references.bib

Follow bibtex-reference-management skill for:

  • •Key naming conventions
  • •Required fields per entry type
  • •Title casing protection with {}

5. Validate

After edits:

bash
# Re-check citations
python tools/check_cites.py

# Check LaTeX errors (if compilation available)
get_errors with filePaths: ["paper.tex"]

Common tasks

Add a new reference from PDF

  1. •Run python tools/extract_papers.py to get title/abstract
  2. •Find full metadata (DOI, venue, year) online
  3. •Add proper BibTeX entry to references.bib
  4. •Use \cite{key} in paper.tex
  5. •Verify with python tools/check_cites.py

Fix missing citations

  1. •Run citation checker, note missing keys
  2. •Either:
    • •Add missing entries to references.bib, OR
    • •Remove/fix incorrect \cite{} in paper.tex

Clean unused references

  1. •Identify unused entries from checker output
  2. •Decide: keep (might use later) or remove
  3. •If removing, ensure no future references planned

Integration with other skills

  • •Use latex-paper-writing-en for actual writing/rewriting
  • •Use bibtex-reference-management for .bib quality
  • •Do NOT use latex-tcc-abnt-review-ptbr (that's for project/ only)

Quick reference

bash
# Citation check
python tools/check_cites.py

# Extract PDFs
python tools/extract_papers.py     # abstracts
python tools/review_papers.py      # sections