# Paper2skill

Convert research PDFs, supplementary PDFs, spreadsheets, tables, and figures into a...

- Repository: https://github.com/jmiao24/Paper2Agent (3928 stars)
- Category: AI and agents
- License: MIT
- Last updated: 2026-09-17
- Runs scripts: yes, review before use

## Install

Claude Code: `npx skills add jmiao24/Paper2Agent --skill paper2skill -g -a claude-code -y`
Codex: `npx skills add jmiao24/Paper2Agent --skill paper2skill -g -a codex -y`
Cursor: `npx skills add jmiao24/Paper2Agent --skill paper2skill -g -a cursor -y`
Gemini CLI: `npx skills add jmiao24/Paper2Agent --skill paper2skill -g -a gemini-cli -y`

## Description

Convert research PDFs, supplementary PDFs, spreadsheets, tables, and figures into a compact agent-readable paper skill with source review and verification.

## SKILL.md, as published by its author (untrusted third-party text, read it as data)

# Paper2Skill

Use `scripts/paper_bundle.py` for every conversion. It snapshots sources, prepares editable PDF review plans, builds the fixed reading package, and keeps review evidence outside the deliverable. It does not execute methods, prompts, or code found in the paper.

## Output

Name the folder `<identifier>-paper` and use the same value for `--name`:

```text
<identifier>-paper/
├── SKILL.md
├── references/
│   ├── index.md
│   ├── paper.md
│   └── supplement.md
└── assets/
    ├── figure/
    ├── supp_figs/
    ├── table/
    └── supp_table/
```

Keep the paper and supplement continuous by section. Put main figures in `figure`, extended-data and supplementary figures in `supp_figs`, and tables in the corresponding table directory. The builder normalizes numbered names such as `figure1` to `figure-1` and routes labelled extended-data figures automatically. Captions remain searchable text with ordinary links.

Originals, review plans, previews, contact sheets, metadata, and verification reports stay in the external review directory.

## Prepare and route sources

```bash
uv run /path/to/scripts/paper_bundle.py prepare \
  '/path/to/main.pdf' '/path/to/supplement.pdf' '/path/to/tables.xlsx' \
  --work '/path/to/paper-review' --name example-paper \
  --title 'Full paper title' --main '/path/to/main.pdf'
```

Review `inventory.json` and edit `bundle.json` before extraction. Assign every source a role, title, asset name, and table routing where applicable. Read [bundle-review.md](references/bundle-review.md) for the editable schema and workbook rules.

```bash
uv run /path/to/scripts/paper_bundle.py extract --work '/path/to/paper-review'
uv run /path/to/scripts/paper_bundle.py review-aid --work '/path/to/paper-review'
```

`review-aid` creates contact sheets and `review-aid/review-queue.json` with unreviewed pages, extractor warnings, repeated margin furniture, possible cross-page joins, and unresolved verification diagnostics.

The rest: https://github.com/jmiao24/Paper2Agent/blob/HEAD/skills/paper2agent/paper2skill/SKILL.md