Skip to content
Get packs

Guide · Agent skill

How to use the PDF agent skill: create, fill and edit PDFs

Updated · written with AI from the project's own documentation and checked against our catalogue data

The pdf skill teaches an AI agent to create and modify PDF files for you: generate a simple report from text, build or fill form fields, merge, split, rotate and watermark PDFs, password-protect them, and extract text. This guide explains what the skill does, how to install it, how to phrase requests so it triggers, and what to check before you share the output.

PDF creation and editingspinabot/brigade · MIT

Quick start

  1. 1Download the ZIP. Download the pdf skill ZIP from this page.
  2. 2Install the skill. Unzip it into ~/.claude/skills/ for personal use, or into a project's .claude/skills/ folder. In the Claude apps, upload it under Settings → Capabilities → Skills.
  3. 3Check the requirements. Confirm your Brigade runtime has the exec-node command. LibreOffice or a headless browser is only needed for HTML-to-PDF layouts.
  4. 4Make a natural request. Describe the PDF task in plain language, naming the input files and the output you want.
  5. 5Review the result. Open the output, check page order and field values, and confirm the file opens before sharing.

What is an agent skill and what does the pdf skill do?

An agent skill is a folder containing a SKILL.md file and any helper files that teach an AI agent how to handle a specific kind of task. The pdf skill is one of these. Once it is available, the agent knows how to produce and change PDF files on your behalf.

The skill covers four kinds of work: generating a content PDF from a title, headings and paragraphs; filling or creating AcroForm fields (text, checkbox, radio and dropdown); editing existing PDFs by merging, splitting, rotating, stamping, watermarking or encrypting them; and drawing precise layouts with custom or CJK/Unicode fonts.

  • Generate simple reports, summaries and letters from text (US-Letter, single column).
  • Fill existing fillable form fields and check the values afterwards.
  • Create new fillable form fields in a PDF.
  • Merge, split, rotate, add or remove pages.
  • Stamp text or add a 45° translucent watermark.
  • Password-encrypt a PDF.
  • Extract text from a PDF.

How do you install the pdf skill?

Download the pdf skill ZIP from this page. Then choose one of two locations. For a personal install, unzip it into ~/.claude/skills/. For a single project, unzip it into the project's .claude/skills/ folder.

If you use the Claude apps, you can instead upload the skill under Settings → Capabilities → Skills.

  • Personal: unzip into ~/.claude/skills/
  • Project only: unzip into .claude/skills/ inside the project folder
  • Claude apps: Settings → Capabilities → Skills, then upload

What does the skill need to run?

The skill needs the Brigade runtime with the exec-node command, because its precise-layout path runs a small JavaScript script. The bundled Node.js packages are @cantoo/pdf-lib, @pdf-lib/fontkit and unpdf, as listed in SKILL.md. LibreOffice (soffice) or a headless browser is optional and only needed for HTML-to-PDF layouts.

The skill runs scripts, but according to this listing it does not use the network.

  • Brigade runtime with the exec-node command
  • Bundled packages: @cantoo/pdf-lib, @pdf-lib/fontkit, unpdf
  • Optional: LibreOffice (soffice) or a headless browser for HTML-to-PDF

How do you trigger the skill with a request?

You do not need a special command. Ask in plain language that names the PDF task. The skill is meant to be used when you ask the agent to make a PDF, fill or build a PDF form, or combine, split, stamp or secure PDFs.

Name the file, say what it should contain, and say where the output should go. The more specific the request, the easier it is to check the result.

  • "Make a one-page PDF summary of this meeting text, titled Q3 Review."
  • "Fill in the name and date fields of contract.pdf with the values from this list."
  • "Merge report-1.pdf and report-2.pdf into one file in that order."
  • "Add a DRAFT watermark to proposal.pdf."
  • "Password-protect payslip.pdf and tell me before you share it."

What should you check before relying on the output?

The skill is designed to verify its own work. It confirms the output file exists and opens, and for filled or created forms it reads back the field values before reporting success. Where layout matters, it can render the page to an image and look at placement.

You should still look at the result yourself. PDF is a final format with no text reflow, so editing the wording of an existing PDF is not reliable. If the recipient needs to change the text, ask for an editable source such as a .docx file alongside the PDF.

  • Open the file and confirm the page count and order.
  • Check every filled field value against your source data.
  • For scanned or flat forms, check that drawn text lines up with the printed boxes.
  • Confirm any password is stored somewhere safe before sending.

What are the safety notes?

The skill can write files and set passwords, and it runs scripts. Treat it like any tool that changes your files: review the output before sharing it, especially for sensitive documents such as contracts, payslips or identity forms.

Scanned or flat forms without real fields are a weak case. The skill then places text by coordinates, and if the layout is wrong the text can land in the wrong place.

  • Review outputs before sharing sensitive documents.
  • Do not assume a scanned form has fillable fields.
  • Keep your original PDFs; the skill writes new output files.
  • The listing says the skill does not use the network, but it does run scripts.

What people use it for

  • Turning meeting notes or a summary into a formatted one-page PDF report.
  • Filling the name, date and amount fields of a fillable form from a list of values.
  • Merging several monthly reports into one file for a client.
  • Adding a DRAFT watermark to a proposal before circulating it.
  • Creating a new fillable form with text and checkbox fields for collecting information.

Frequently asked questions

How do I install the pdf agent skill?

Download the ZIP from this page and unzip it into ~/.claude/skills/ for personal use, or into a project's .claude/skills/ folder. In the Claude apps, you can upload it under Settings → Capabilities → Skills instead.

What do I need to run the pdf skill?

You need the Brigade runtime with the exec-node command. The bundled packages are @cantoo/pdf-lib, @pdf-lib/fontkit and unpdf. LibreOffice or a headless browser is optional and only needed for HTML-to-PDF layouts.

Is the pdf skill free to use?

The skill is published under the MIT licence. This page does not list pricing, so check the directory listing for any cost details.

Can the pdf skill edit the text of an existing PDF?

Not reliably. PDF has no reflow engine, so changing flowing text in place is not dependable. The recommended approach is to keep an editable source file such as .docx and deliver the PDF generated from it.

Can it fill a scanned PDF form?

Only if you accept drawn text placed by coordinates. Scanned or flat forms have no real fields, so text can misalign if the layout is wrong. Check the placement carefully.

Does the pdf skill send my files over the internet?

The listing states that the skill does not use the network. It does run scripts and write files locally, so review outputs and treat sensitive documents with care.

New agent skills, repos and AI tools — every week

The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.

Our verdict

Use the pdf skill if you need quick formatted PDFs from text, fillable forms completed with checked values, or routine merging, splitting and watermarking. Pick another approach if you need to rewrite the wording of an existing PDF or work on scanned forms without fields, where an editable source file is the better route.

Download the skill