How to Search a PDF From Your Editor
Every integration has one: the partner spec, the compliance manual, the vendor's 80-page PDF that governs what your code must do. Your questions about it arrive mid-function — and the answer ritual is alt-tab, scroll, squint, Ctrl+F, lose your place twice.
With Apdf connected to Claude Code or Cursor over MCP, the spec becomes something you ask from the editor — and the answers come back with the exact line and page number, ready to paste into a code comment.
Connect your editor to Apdf
In Claude Code it's one command; Cursor and other MCP clients take the same endpoint in their MCP settings:
claude mcp add --transport http apdf https://apdf.io/mcp/mainThe spec just needs a reachable URL — your team drive, object storage, or a private Apdf document.
Ask mid-function
https://files.northlight.example/specs/paylane-spec.pdf
Under the answer, search-pdf-content returned the
receipts — the exact line, with its page:
{
"search_text": "backoff",
"results_total": 1,
"results_pages": 1,
"results": [
{
"page": "3",
"matched_line": "Retries: 5 attempts with exponential backoff starting at 30 seconds.",
"exact_word": "backoff"
}
]
}Go regex when the question is a pattern
The search accepts regular expressions — which turns “list every error code the spec defines” into one call:
{
"search_text": "ERR-\\d+",
"results_total": 3,
"results_pages": 1,
"results": [
{
"page": "4",
"matched_line": "Errors use stable codes: ERR-1401 expired token, ERR-2201 duplicate idempotency key,",
"exact_word": "ERR-1401"
},
{
"page": "4",
"matched_line": "Errors use stable codes: ERR-1401 expired token, ERR-2201 duplicate idempotency key,",
"exact_word": "ERR-2201"
},
{
"page": "4",
"matched_line": "ERR-3305 payout account frozen. Codes never change meaning between versions.",
"exact_word": "ERR-3305"
}
]
}read-pdf-content returns whole
pages for context, and for scanned specs the
search-pdf-content-with-ocr variant runs an OCR pass first —
the OCR tutorial
covers it.
Where to go from here
The same connection carries every PDF operation — and the engagement layer behind them.
Related tutorials
Building a "Search Inside PDF" Feature for Your Web App
This tutorial shows how to build a PDF search feature using Python and the Apdf Search API. Find specific terms, get page numbers and context lines, and optionally use regex patterns for advanced matching. Includes a Flask web interface example.
Turn Scanned Invoices into Structured Data with an Agent
OCR reads tabular scans in scrambled order — the wall every regex parser dies against, and exactly where the agent shines: it reassembles vendor, number, date and line items from the mess, cross-checks the sums against each stated total, and flags mismatches for human eyes instead of your books.
Watermark Draft PDFs via Claude Before Sharing
The draft that leaks is the one you didn't stamp. Conversationally it's one message: the agent designs the translucent DRAFT — CONFIDENTIAL stamp itself with the PDF generator, overlays it across every page, renders a page to prove it looks right — and mints one named copy per reviewer.
Your code made the PDF.
Then it went dark.
Opened, read, re-read, dropped on page 4 — you never see any of it. Share the PDFs you generate through Apdf recipient links, and every signal becomes something you can act on: ping Slack, update the CRM, let an agent follow up. Same account, same API token, one more call.
API · Webhook · MCP