| name | pdf-analysis |
| description | Analyze PDF documents including reports, whitepapers, ebooks, and manuals. Use when the user mentions PDF, document, whitepaper, ebook, report file, or wants to analyze a .pdf file. |
PDF Document Analysis
Analyze PDF documents to generate structured reports with summaries, key points, data extraction, and insights.
When to Use
Activate this skill when the user:
- Mentions "PDF", "document", "whitepaper", "ebook", "report"
- Provides a file path to a .pdf file
- Asks to summarize or analyze a PDF document
- Wants to extract insights from a document file
- References a downloaded report or paper
Instructions
- Get the file path - Ask the user for the PDF file path if not provided
- Verify the file exists and is readable
- Read the PDF content using the Read tool
- The Read tool can extract text from PDF files
- If reading fails:
- If PDF appears to be scanned/image-based: Inform user OCR may be needed
- Suggest alternative approaches if text extraction fails
- Read the analysis prompt from
prompts/pdf.md - Extract document metadata from content:
- Title (from content or derive from filename)
- Author if mentioned
- Date if mentioned
- Generate analysis following the prompt structure exactly
- Create output directory
reports/pdfs/if needed - Save the report to
reports/pdfs/YYYY-MM-DD_sanitized-title.mdwhere:- YYYY-MM-DD is today's date
- sanitized-title is the title in lowercase, spaces replaced with hyphens, special chars removed, max 50 chars
- Update the activity log at
logs/YYYY-MM-DD.md:- Create file if it doesn't exist with standard sections
- Add entry under "## PDFs Reviewed" section (create if needed)
- Format:
- [Title](../reports/pdfs/filename.md) - HH:MM
- Confirm to user what was saved and where
Report Format
Include this header in the report:
# [Document Title]
**Source**: [file path]
**Date**: YYYY-MM-DD
**Type**: PDF Document
---
[Analysis content following prompts/pdf.md structure]
---
## My Notes
[Empty space for user notes]
Error Handling
- If file path doesn't exist: Ask user for correct path
- If file is not a PDF: Suggest using /analyze for other file types
- If text extraction fails: Suggest OCR or manual copy
- If prompts/pdf.md missing: Use prompts/default.md
Related
- Slash command equivalent:
/pdf <filepath> - Prompt file:
prompts/pdf.md - Output location:
reports/pdfs/