paper-downloader

# Paper Downloader Skill ## Purpose Download paper PDFs and clone associated source code repositories. This is the second step in the paper reproduction pipeline, following paper-finder. ## When to Use - After running paper-finder to get paper_info.json - User asks to "download", "get", or "fetch" a paper's PDF - User asks to "clone" or "get the code" for a paper - User asks to "reproduce" a paper (run paper-finder first, then this) ## How to Use ### Step 1: Ensure paper_info.json exists Run paper-finder first if needed: ```bash python skills/paper-finder/scripts/search_paper.py "<query>" --output workspace/paper_info.json ``` ### Step 2: Run the Download Script ```bash python skills/paper-downloader/download_paper.py workspace/paper_info.json ``` **Options:** - `--allow-scihub` — Enable Sci-Hub as fallback PDF source - `--clone-depth 0` — Full git clone (default: shallow clone) - `--skip-code` — Only download PDF - `--skip-pdf` — Only clone code - `-o <dir>` — Custom output directory ### Step 3: Read the Results Check `workspace/<paper-slug>/download_report.json` for: - PDF download status and path - Code clone status and path - Step-by-step log ### Step 4: Report to User Tell the user: - Where the PDF was saved - Where the code was cloned - Which dependency files were found in the repo ## Dependencies - Python 3.10+ - `requests` library - `git` command-line tool