Document Processing and Extraction Agent

# Document Processing and Extraction Agent

You are a document automation engineer. Design a processing agent for [DOCUMENT_TYPE].

## Document Context

- Document category: [DOCUMENT_TYPE] (e.g., contracts, invoices, reports, applications)
- Volume: approximately [VOLUME] documents per [PERIOD]
- Input format: [INPUT_FORMAT] (PDF, Word, email, scanned image)
- Current manual process: [CURRENT_PROCESS]

## Extraction Schema

Define what the agent extracts from each document:

For each field, specify:
- Field name
- Data type
- Where it typically appears in the document
- How to handle missing or ambiguous values
- Validation rule

Fields to extract: [FIELD_LIST]

## Processing Pipeline

Design the 5-stage pipeline:
1. Ingestion: how documents enter the system
2. Parsing: how text is extracted from raw files
3. Extraction: how fields are identified and pulled
4. Validation: how errors and edge cases are caught
5. Output: where results go and in what format

## Edge Case Handling

Write handling logic for:
- Document is missing a required field
- Field value is ambiguous
- Document format is unrecognized
- Extraction confidence is below [CONFIDENCE_THRESHOLD]%

## Output Schema

Show the exact JSON structure each processed document produces.