Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
Version History
Fix doc/code mismatch: convert.sh now uses uvx markitdown[pdf]; download.sh now includes Sci-Hub as Strategy 5; add legal disclaimer for Sci-Hub in SKILL.md
What This Skill Does
Manages academic literature collections end-to-end: search Google Scholar, download PDFs by DOI, convert to markdown, organize into a structured folder hierarchy with JSON and human-readable indexes, and audit for broken paths or inconsistencies. Handles paywalled papers via open-access sources and Sci-Hub fallback.
Covers the full search-to-verified-collection pipeline in one skill rather than requiring separate tools for each stage.
When to use it
- Building a categorized PDF library for a literature review
- Downloading a list of DOIs and converting all papers to searchable markdown
- Auditing an existing references folder for missing files or stale paths
- Collecting GitHub repos and datasets cited across a set of method papers
- Incrementally adding newly found papers to an organized existing collection
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: Find and download 10 recent papers on protein language models, organized by topic
- 1Searches Google Scholar with a date filter to find relevant papers and extracts titles, authors, DOIs, and PDF links
- 2Runs scripts/download.sh for each DOI, trying open-access sources then Sci-Hub for paywalled papers
- 3Runs scripts/convert.sh to produce a markdown file for each downloaded PDF
- 4Creates the references/ folder structure with category subfolders, writes index.json and README.md
- 5Runs scripts/audit.sh to verify every PDF is valid, has a matching markdown file, and all index.json paths resolve
A references/ directory with categorized PDFs, markdown files, index.json, README.md, and a clean audit report
Requirements
Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.