Paper Audit

Unified paper audit for Chinese and English papers. Use when reviewing paper quality, pre-submission checks, or doing adversarial reviews.

jsgforever fcc29a6 16 files · 125.9 KB Updated

File contents

Paper Audit Skill

Unified academic paper auditing across formats and languages.

Steps

  1. Parse the paper path and mode from $ARGUMENTS. If missing, confirm the target .tex, .typ, or .pdf file.
  2. Review evaluation criteria in $SKILL_DIR/references/REVIEW_CRITERIA.md and $SKILL_DIR/references/CHECKLIST.md.
  3. Run the orchestrator: python $SKILL_DIR/scripts/audit.py $ARGUMENTS.
  4. Present the MD report directly to the user. Distinguish between automated findings and LLM-judgment scores.
  5. If in review mode, read $SKILL_DIR/references/SCHOLAR_EVAL_GUIDE.md and formulate LLM assessments for Novelty, Significance, Ethics, and Reproducibility.
  6. If in polish mode, read .polish-state/precheck.json and spawn nested tasks for the Critic Agent and Mentor Agents as defined in the original workflow specifications.

Examples

  • Run python $SKILL_DIR/scripts/audit.py $ARGUMENTS.
  • Run python $SKILL_DIR/scripts/audit.py paper.pdf --mode review.

Troubleshooting

  • Provide a valid .tex, .typ, or .pdf path when arguments are missing.
  • Report command output and exit code when audit execution fails.

jsgforever/skillguard/tree/main/benchmarks/final/skills/asset1__bahayonghang--paper-audit commit fcc29a608a

Frequently asked questions

npx skillmds@latest add jsgforever/paper-audit