PDF Skill

Use when the task involves creating, editing, reviewing, splitting, merging, or extracting content from PDF files.
Use when the user needs to understand whether a PDF can be edited directly or must be converted first.
Use when the task depends on preserving layout, images, pagination, or annotation fidelity.
Use when OCR may be required because the PDF contains scanned pages rather than selectable text.

Chooses the safest PDF workflow based on the user's goal: review, extraction, transformation, or generation.
Distinguishes between text-based PDFs and image-only scanned PDFs before proposing edits.
Preserves fidelity when layout matters and explicitly calls out any tradeoff when converting to another format.
Prefers repeatable tools and scripted operations over manual one-off desktop steps.

Start by identifying the real goal: review content, extract data, annotate, merge, split, redact, or convert.
Determine whether the PDF is text-based or scanned. If the content is image-only, state that OCR may be required.
If the user needs edits that preserve layout, avoid lossy conversions unless they explicitly accept them.
If extracting text, note whether the output should preserve reading order, tables, headings, or page boundaries.
If generating a PDF from another source, preserve the source of truth in an editable format whenever possible.
When tools are limited, explain the limitation clearly and offer the safest fallback rather than implying full fidelity.
Summarize important tradeoffs before carrying out destructive or lossy transformations.

Never imply that all PDFs are safely editable; many require conversion or specialized tooling.
Call out OCR accuracy risk explicitly for scanned documents.
Preserve the original file before transformations that may alter layout or metadata.
Prefer deterministic, scriptable workflows for extraction and generation.
Tell the user when tables, forms, annotations, or embedded fonts may not round-trip cleanly.

PDF Skill

When to use this skill

Use when the task involves creating, editing, reviewing, splitting, merging, or extracting content from PDF files.

Use when the user needs to understand whether a PDF can be edited directly or must be converted first.

Use when the task depends on preserving layout, images, pagination, or annotation fidelity.

Use when OCR may be required because the PDF contains scanned pages rather than selectable text.

What this skill does

Chooses the safest PDF workflow based on the user's goal: review, extraction, transformation, or generation.

Distinguishes between text-based PDFs and image-only scanned PDFs before proposing edits.

Preserves fidelity when layout matters and explicitly calls out any tradeoff when converting to another format.

Prefers repeatable tools and scripted operations over manual one-off desktop steps.

Implementation

Start by identifying the real goal: review content, extract data, annotate, merge, split, redact, or convert.

Determine whether the PDF is text-based or scanned. If the content is image-only, state that OCR may be required.

If the user needs edits that preserve layout, avoid lossy conversions unless they explicitly accept them.

If extracting text, note whether the output should preserve reading order, tables, headings, or page boundaries.

If generating a PDF from another source, preserve the source of truth in an editable format whenever possible.

When tools are limited, explain the limitation clearly and offer the safest fallback rather than implying full fidelity.

Summarize important tradeoffs before carrying out destructive or lossy transformations.

Best practices

Never imply that all PDFs are safely editable; many require conversion or specialized tooling.

Call out OCR accuracy risk explicitly for scanned documents.

Preserve the original file before transformations that may alter layout or metadata.

Prefer deterministic, scriptable workflows for extraction and generation.

Tell the user when tables, forms, annotations, or embedded fonts may not round-trip cleanly.

pdf