How do I install the PDF Viewer skill?

Run npx skillmds@latest add exodes/pdf-viewer in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is PDF Viewer safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with PDF Viewer?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is PDF Viewer free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published PDF Viewer?

eXodes (@exodes) published this skill. Their other Agent Skills are listed on their SkillMD profile.

PDF Viewer by exodes — Agent Skill

Name: PDF Viewer
Author: eXodes

PDF Viewer

Interactive PDF viewer skill. Use when user wants to open, view, annotate, highlight, stamp, fill form fields, or sign a PDF with live visual feedback. Trigger phrases: "open this PDF", "show me this contract", "highlight the key terms", "help me fill out this form", "sign this", "stamp it APPROVED", "mark this CONFIDENTIAL", "add initials to each page", "walk me through this document". Also trigger when user provides a PDF path or arXiv URL alongside any visual/interactive intent. Do NOT trigger for pure summarization or text extraction (use native Read tool instead).

eXodes 9ce7a4a 6 files · 17.0 KB Updated Sep 17, 2026

File contents

pdf-viewer

You have access to a local PDF server that renders documents in a live viewer and supports annotation, form-filling, and signature placement with real-time visual feedback.

Reference files

Load on demand — do not load all eagerly:

File	What's in it	When to read
`references/view-pdf.md`	Full tool API (`display_pdf`, `interact`), annotation types, coordinate system	Always read on activation — core reference
`references/open.md`	Opening workflow, supported sources, when NOT to use viewer	When user just wants to open/view a document
`references/annotate.md`	Collaborative annotation workflow, batch patterns, tips	When user wants to annotate, highlight, stamp, or mark up
`references/fill-form.md`	Form-filling workflows (user-driven vs AI-assisted), unnamed field handling	When user wants to fill a form
`references/sign.md`	Signature/initials placement workflow, coordinate tips	When user wants to sign or add initials

Quick decision guide

User intent	Action
"open" / "show" / "view"	`display_pdf` → offer next steps based on doc type
"annotate" / "highlight" / "stamp" / "mark up"	Read `annotate.md`
"fill" / "form" / "complete this form"	Read `fill-form.md`
"sign" / "initials" / "signature"	Read `sign.md`
"summarize" / "what does it say" / "extract"	Stop — use native Read tool, not viewer

Core tools (summary)

list_pdfs — list available local PDFs. Call when no path given.
display_pdf(url, page?, elicit_form_inputs?) — open document. Returns viewUUID (required for all interact calls) and formFields if PDF has fillable fields.
interact(viewUUID, commands[]) — all follow-up actions. Batch commands; end with get_screenshot to verify.

Full API detail in references/view-pdf.md.

Always

Call display_pdf once per document. Reuse viewUUID for all interact calls.
Batch related commands in one interact call.
End annotation/form/sign batches with get_screenshot — show result to user.
Remind user they can download annotated PDF from viewer toolbar when done.

---name: pdf-viewerdescription: Interactive PDF viewer skill. Use when user wants to open, view, annotate, highlight, stamp, fill form fields, or sign a PDF with live visual feedback. Trigger phrases: "open this PDF", "show me this contract", "highlight the key terms", "help me fill out this form", "sign this", "stamp it APPROVED", "mark this CONFIDENTIAL", "add initials to each page", "walk me through this document". Also trigger when user provides a PDF path or arXiv URL alongside any visual/interactive intent. Do NOT trigger for pure summarization or text extraction (use native Read tool instead).---# pdf-viewerYou have access to a local PDF server that renders documents in a live viewerand supports annotation, form-filling, and signature placement with real-timevisual feedback.## Reference filesLoad on demand — do not load all eagerly:| File | What's in it | When to read || ------------------------- | ------------------------------------------------------------------------------ | --------------------------------------------------------- || `references/view-pdf.md` | Full tool API (`display_pdf`, `interact`), annotation types, coordinate system | Always read on activation — core reference || `references/open.md` | Opening workflow, supported sources, when NOT to use viewer | When user just wants to open/view a document || `references/annotate.md` | Collaborative annotation workflow, batch patterns, tips | When user wants to annotate, highlight, stamp, or mark up || `references/fill-form.md` | Form-filling workflows (user-driven vs AI-assisted), unnamed field handling | When user wants to fill a form || `references/sign.md` | Signature/initials placement workflow, coordinate tips | When user wants to sign or add initials |## Quick decision guide| User intent | Action || ---------------------------------------------- | -------------------------------------------------- || "open" / "show" / "view" | `display_pdf` → offer next steps based on doc type || "annotate" / "highlight" / "stamp" / "mark up" | Read `annotate.md` || "fill" / "form" / "complete this form" | Read `fill-form.md` || "sign" / "initials" / "signature" | Read `sign.md` || "summarize" / "what does it say" / "extract" | **Stop** — use native Read tool, not viewer |## Core tools (summary)- **`list_pdfs`** — list available local PDFs. Call when no path given.- **`display_pdf(url, page?, elicit_form_inputs?)`** — open document. Returns `viewUUID` (required for all `interact` calls) and `formFields` if PDF has fillable fields.- **`interact(viewUUID, commands[])`** — all follow-up actions. Batch commands; end with `get_screenshot` to verify.Full API detail in `references/view-pdf.md`.## Always- Call `display_pdf` once per document. Reuse `viewUUID` for all `interact` calls.- Batch related commands in one `interact` call.- End annotation/form/sign batches with `get_screenshot` — show result to user.- Remind user they can download annotated PDF from viewer toolbar when done.

Frequently asked questions