| name | download-gated-pdfs |
| description | Download the actual PDF binary from bot-gated sites (taxpolicycenter.org, urban.org,
SSRN-hosted mirrors, think-tank/publisher sites) via the Wayback Machine id_ URL form.
Use when: (1) curl/WebFetch of a .pdf URL returns HTML instead of a PDF even with a
browser User-Agent, (2) pypdf fails with "invalid pdf header: b'<!DOC'" or
"EOF marker not found" on a freshly downloaded file, (3) Firecrawl can parse the PDF
to markdown but you need the original file on disk (e.g., filing a reference copy).
|
| author | Claude Code |
| version | 1.0.0 |
| date | 2026-07-15T00:00:00.000Z |
Download bot-gated PDFs via Wayback id_
Problem
Many think-tank and publisher sites (taxpolicycenter.org, urban.org, SSRN delivery)
serve an HTML bot-challenge page instead of the PDF to non-browser clients. A browser
User-Agent header does not help. The downloaded "PDF" is actually HTML.
Context / Trigger Conditions
curl -o file.pdf <url> succeeds but the file starts with <!DOC- pypdf raises
invalid pdf header: b'<!DOC' or PdfStreamError: Stream has ended unexpectedly
- Firecrawl
scrape returns clean markdown for the same URL (its proxies get through),
but Firecrawl does not return the binary — only parsed content
Solution
- Request the file through the Wayback Machine's raw-content (
id_) endpoint, which
serves the original archived binary without rewriting:
curl -sL -A "Mozilla/5.0 ... Chrome/126.0 Safari/537.36" \
"https://web.archive.org/web/<YYYY>id_/<original-pdf-url>" -o out.pdf
<YYYY> is any year likely to have a snapshot (e.g. publication year); Wayback
redirects to the nearest capture. The id_ suffix after the timestamp is what
requests the untouched original.
- Verify the download with pypdf — a bot page fails immediately:
from pypdf import PdfReader
r = PdfReader("out.pdf"); print(len(r.pages), "pages")
- If Wayback has no capture, fall back to: another mirror found via search
(Exa/Firecrawl), or Firecrawl scrape for the parsed text when the binary is not
strictly needed.
Verification
PdfReader opens the file and reports a plausible page count; first-page text matches
the expected title.
Example
Verified 2026-07-15: taxpolicycenter.org/sites/default/files/publication/165884/ssrn-id4797771.pdf
and urban.org/sites/default/files/publication/80621/2000790-...pdf both bot-gated to
direct curl (with UA), both downloaded intact via
https://web.archive.org/web/2024id_/<url> and .../web/2023id_/<url> (18 and 12 pages).
Notes
- Government data hosts (e.g.
ticdata.treasury.gov) are usually NOT gated — try direct
curl first; Wayback is the fallback, not the default.
- Wayback captures can be stale for frequently-revised documents; check the snapshot date
if currency matters.
- See also: the
pdf skill (parsing/extraction after download) and
firecrawl:firecrawl-scrape (parsed markdown when the binary is unnecessary).
