Extract Requirements

How to use me

Call me with a PDF file path:

Load the extract-requirements skill and analyze this PDF: /path/to/requirements.pdf

The skill will:

1. Extract the repository name from your current git repository
2. Read and analyze the PDF file
3. Identify main requirements and associated sub-tasks
4. Generate a requirements.json file in the appropriate Reports directory
5. Output the results and save location

Output Format

The generated requirements.json follows this structure:

{
  "repository_name": "project-name",
  "extraction_date": "2026-03-18",
  "total_requirements": 5,
  "main_requirements": [
    {
      "id": "REQ-001",
      "title": "Requirement Title",
      "description": "Detailed description of the requirement",
      "sub_tasks": [
        {
          "id": "TASK-001",
          "title": "Sub-task Title",
          "description": "Description of the sub-task"
        }
      ]
    }
  ]
}

Key Features

• Semantic Analysis - Understands requirement context, not just keywords
• Flexible Structure - Works with unstructured PDF documents
• Automatic Organization - Hierarchically organizes requirements and tasks
• Git Integration - Automatically determines repository name from git
• Structured Output - Clean JSON format for downstream tools

Requirements

• Repository must be a git repository
• PDF file must be readable and contain structured content
• Write access to /Users/admin/dev/Reports/{repository_name}/