- Home
- Skills
- APIs & Backend
- PDF Reading
PDF Reading
Process, generate, and verify PDFs with visual rendering and reliable AcroForm handling.
Free
Works with the AI Agents you already use
PDF Reading
Example session with this skill installed
I need to fill the attached 'tax_form.pdf' with my name 'Jane Doe' and total '5000'. Keep the form interactive and verify the fields are correctly mapped.
- Read your context and instructions
- Compiled the pdf reading
I have updated the interactive fields in the tax form. I performed an AcroForm repair to ensure all widgets were correctly mapped to their parent fields, filled the values, and verified the final rendering to ensure no text was clipped.
PDF Form Update
| Action | Detail | Result |
|---|---|---|
| Field Mapping | full_name -> "Jane Doe" | Success |
| Field Mapping | total_amount -> "5000" | Success |
| Widget Repair | reattach_fields() | 2 orphaned widgets recovered |
| Visual Check | pdftoppm render | Layout verified, no clipping |
# Verification Snippet
from pypdf import PdfReader
reader = PdfReader("output/pdf/tax_form_filled.pdf")
fields = reader.get_fields()
print(fields['full_name']['/V']) # Output: Jane Doe
Modified :codex-file-citation{path="output/pdf/tax_form_filled.pdf" purpose="output"}, with the updated interactive fields and verified visual layout.
Next steps
- :codex-followup[Flatten the form]{prompt="Flatten tax_form_filled.pdf into a static document for submission."}
- :codex-followup[Add a digital signature]{prompt="Help me add a signature field to the bottom of the first page."}
- :codex-followup[Extract all field values]{prompt="Export all current field values from tax_form_filled.pdf to a JSON file."}
Connects securely to your tools. The creator never sees your data.
About this skill
Processing PDF files often results in broken layouts, lost form data, or unreadable text extraction. This skill provides a systematic approach to reading, creating, and verifying PDFs where visual fidelity is mandatory. It moves beyond simple text scraping by using visual rendering to confirm that what the code sees matches what the user will see.
What it does
- Visual verification renders PDF pages to PNGs using Poppler to catch layout shifts, clipped text, or overlapping elements.
- Form management handles AcroForms by recovering orphaned widgets and ensuring logical field data matches visual appearances.
- Programmatic generation creates documents with precise typography and margins using
reportlab. - Hybrid extraction combines
pdfplumberandpypdfto pull data while maintaining awareness of document structure.
How it works
- Inspection uses
pdftoppmto generate high-resolution images of the document for visual alignment checks. - Form handling runs a recovery workflow to reattach fields and widgets before updating interactive values.
- Artifact creation executes Python scripts to build or modify documents, followed by a mandatory re-rendering step.
- Validation compares the logical data in the PDF tree against the final rendered output to ensure consistency.
Frameworks & tools
This skill utilizes Python libraries including reportlab, pdfplumber, and pypdf. It relies on system-level Poppler utilities (pdftoppm, pdfinfo) for rendering and metadata inspection.
Why this beats prompting it yourself
Standard prompts often fail at the nuances of the PDF specification, such as the discrepancy between /AcroForm fields and page /Annots. This skill enforces a verification loop that renders the file to confirm visual correctness, preventing the delivery of broken or clipped documents.
Use cases
- Generating invoices or reports that require strict brand alignment and margin control.
- Filling complex government or corporate forms while keeping them interactive.
- Auditing existing PDFs for rendering defects or inconsistent field metadata.
- Extracting structured data from documents where visual layout is the only reliable guide.
Known limitations
Requires Poppler to be installed on the host system for visual rendering. Text extraction does not guarantee 100% layout fidelity without visual verification.
Requires Poppler (pdftoppm) for visual rendering. Form flattening removes all interactivity. Text extraction may not reflect complex visual layouts perfectly without manual review.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
1 use
Times used to date
Free forever
No account required to browse
Trust & safety
Security scan
Security scan passed 1 month ago
- Free with an Agensi account
Requires