Menu
Organize a paper PDF and its codebase into clean reading artifacts. Use this before any other read-* skills.
Installer avec Codex ou Claude Copiez ce prompt, collez-le dans Codex, Claude ou un autre assistant, puis laissez-le vérifier la page du skill et l'installer pour vous.
Basé sur la classification professionnelle SOC
Verify each cited reference exists via web search.
Deep extraction of all verifiable claims from a research paper. Actively proposes implicit claims, splits compound statements, and flags cross-reference inconsistencies.
Verify paper claims with evidence_type "code" against the actual codebase. Use after read-txt to check if code matches what the paper says.
Review experiment design and produce a prioritized reproduction plan. Targets claims with evidence_type "experiment" from check_claim.json. Output feeds into run-exp.
Web-search validation of citation and novelty claims from check_claim.json. Checks whether cited facts are accurate and novelty claims ("first", "pioneering") hold against published literature.
Verify theoretical claims — mathematical derivations, proofs, and formal arguments. Targets claims with evidence_type "theoretical" from check_claim.json.
| name | prepare-paper |
| description | Organize a paper PDF and its codebase into clean reading artifacts. Use this before any other read-* skills. |
| author | PaperDoctor Research |
| license | MIT |
| argument-hint | paper.pdf |
| allowed-tools | Read, Bash(python *), Bash(kill *) |
Use this workflow to organize a paper PDF and code into clean reading artifacts.
+---------------------------+ +---------------------------+ +---------------------------+
| Step 1: Parse PDF | | Step 2: Render Pages | | Step 3: Index Codebase |
| api_mathpix.py | | render page PNGs | | build code index |
+---------------------------+ +---------------------------+ +---------------------------+
\ | /
\ | /
\_________________________|_________________________/
|
v
+---------------------------+
| Step 4: Organize Paper |
| full.md / refs / sections|
+---------------------------+
Parallel rule:
Prefer api_mathpix.py (Mathpix API) first — it produces high-quality markdown with LaTeX math and downloads images locally.
Primary command (try first):
python tools/api_mathpix.py papers/<paper_dir>/<paper_file>.pdf
This uploads the PDF, waits for processing, downloads markdown + images to papers/<paper_dir>/metadata/<arxiv_id>/mathpix/.
Fallback (only if Mathpix API fails — account disabled, network error, etc.):
python tools/api_mineru.py -p papers/<paper_dir>/<paper_file>.pdf -o papers/<paper_dir>/metadata
Command:
python tools/pdf_render.py papers/<paper_dir>/<paper_file>.pdf
Command:
python tools/code_analyzer.py papers/<paper_dir>/<repo_dir> --output papers/<paper_dir>/metadata/code/index.json
Command:
python tools/organize_paper.py --paper-file papers/<paper_dir>/<paper_file>.pdf
This step reads the Mathpix/MinerU output from Step 1 and writes the normalized paper text, extracted references, and per-section markdown files.
Run the workflow like this:
python tools/api_mathpix.py first. Only fall back to api_mineru.py if the Mathpix API fails.{paper_dir}/metadata/{arxiv_id}/mathpix/{arxiv_id}.md by downstream skills.organize_paper.py only after Mathpix/MinerU output exists.metadata/<pdf_stem>/.