# Download Pmc

> Use when bulk-downloading PubMed Central OA tarballs across the standard PMC sections with the EDirect helper script.

- Skill: `vimalinx/download-pmc` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds add vimalinx/download-pmc`
- Raw SKILL.md: https://api.skillmd.com/api/skills/vimalinx/download-pmc/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: vimalinx (https://skillmd.com/u/vimalinx)
- Updated: 2026-09-09
- Page: https://skillmd.com/skills/vimalinx/download-pmc

---


# download-pmc

## Quick Start

- **Command**: `PATH=/home/vimalinx/miniforge3/envs/bio/bin:$PATH /home/vimalinx/miniforge3/envs/bio/bin/download-pmc [-ftp|-https]`
- **Local executable**: `/home/vimalinx/miniforge3/envs/bio/bin/download-pmc`
- **Full reference**: See [references/help.md](references/help.md)

## When To Use This Tool

- Mirror the PMC OA bulk XML archives in batch.
- Pull baseline and incremental tarballs from `oa_comm`, `oa_noncomm`, and `oa_other`.
- Retry failed downloads and verify tarball contents before keeping them.

## Common Patterns

```bash
# 1) Use the default FTP-backed download mode
PATH=/home/vimalinx/miniforge3/envs/bio/bin:$PATH \
  /home/vimalinx/miniforge3/envs/bio/bin/download-pmc
```

```bash
# 2) Force HTTPS retrieval instead of FTP-style listing
PATH=/home/vimalinx/miniforge3/envs/bio/bin:$PATH \
  /home/vimalinx/miniforge3/envs/bio/bin/download-pmc -https
```

## Recommended Workflow

1. Run it from a directory where bulk PMC tarballs are supposed to accumulate.
2. Pick `-ftp` or `-https` deliberately; FTP-style mode is the default.
3. Let the script finish its full pass over `baseline` and `incr` tarballs in all three OA sections.
4. Review stderr for retries or invalid-content deletions before treating the mirror as complete.

## Guardrails

- This is a bulk operational downloader, not a one-file fetcher and not a safe probe.
- The script iterates all of `oa_comm`, `oa_noncomm`, and `oa_other` for both `baseline` and `incr`; there is no fine-grained selector for one archive family.
- It depends on `nquire`, `xtract`, and `skip-if-file-exists` being available on `PATH`.
- Files that download empty or fail XML verification are deleted and retried; a missing file after the run can reflect repeated validation failure, not just an interrupted transfer.