agentleFS
jxtsejxtse/auto-paper-harvester/.claude/skills/paper-download/SKILL.md

paper-download

Skill32 starsChanged 32 days agoMITView on GitHub

Use when the user wants full-text PDFs for papers they already identified: one or more DOIs, doi.org links, or a reference list ("下载这几篇论文的 PDF", "把这个 DOI 列表下下来", "download the PDFs for these references"). Tries open access first, then the publisher's TDM API if the user has a key, then (opt-in) the user's own logged-in Chrome. Not for finding papers, Zotero import, or bypassing paywalls.

  • Reads credentials

1 other project uses a version of this.

Adopted by

Most recently taken up by

  • Other versions1 project uses a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 1

    1. tsingyuai/scientify★ 2.2ksame name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 0 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Add --browser extension only if doctor line 3 is OK and the user agreed to use their Chrome.

    line 45

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

paper-download

Command: auto-paper-download (from https://github.com/jxtse/auto-paper-harvester).

0. Install once

auto-paper-download --help >/dev/null 2>&1 || \
  uv tool install git+https://github.com/jxtse/auto-paper-harvester   # or: pipx install git+...
auto-paper-download doctor

doctor prints three lines, one per download step:

  1. Open access (OpenAlex / Unpaywall / Crossref). Needs an e-mail address, sent to those APIs as a contact. If the line starts with --, ask the user which e-mail to use, then write it to ~/.auto-paper-download/.env as OPENALEX_MAILTO=, UNPAYWALL_EMAIL= and CROSSREF_MAILTO= (one per line). Do not invent an address.
  2. Publisher TDM APIs (Elsevier / Wiley / Springer). Optional. Without a key that step is skipped. Never ask the user to paste keys into the chat; suggest auto-paper-download ui, which stores them locally.
  3. Browser (the user's own Chrome via the Playwright Extension). Optional and manual: the user installs the extension, logs in to their library, and pastes the token in auto-paper-download ui. Only use it if doctor shows it ready.

1. Download

auto-paper-download get 10.1016/j.jocs.2020.101171 10.5194/npg-20-705-2013 -o papers --json
auto-paper-download get --file refs.txt -o papers --json          # any text with DOIs in it
auto-paper-download get --file refs.txt -o papers --json --browser extension   # step 3 on
  • Tell the user first: how many DOIs, where the files go, whether step 3 is on.
  • Add --browser extension only if doctor line 3 is OK and the user agreed to use their Chrome. It opens one tab. If a CAPTCHA does not clear by itself the paper is skipped and the site cools down for 24 h. Never solve a CAPTCHA for the user.
  • Daily caps and per-site pacing are on by default (ledger at ~/.auto-paper-download/ledger.sqlite). Do not pass --no-ledger or raise limits unless the user asks. A rate_limited or host_limited reason means "try again tomorrow".
  • Download only papers the user selected. If the list has more than ~50 DOIs, confirm first.

2. Report

--json prints {"requested", "pdfs", "failed": [{"doi", "reason"}], "output_dir"}. Report: <pdfs>/<requested> saved to <output_dir>, then each failed DOI with its reason in plain words:

reason containstell the user
no_source / all sources DownloadErrorno open copy found; a TDM key or step 3 may get it
rate_limited / host_limited / cooldowndaily limit or a site cooling down; retry later
challengethe site showed a bot check; it was not solved
auth_redirecttheir browser is not logged in to the library

Files: <output_dir>/<doi_slug>/<doi_slug>.pdf plus any supplementary PDFs next to it. Existing files are skipped on re-runs.

More: auto-paper-download get -h, and the repo README.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.