Blocks Reference

RAPFlow provides a comprehensive set of blocks for building AI-powered workflows. Each block serves a specific purpose and can be configured to meet your exact requirements.

Block Categories

Blocks are organized into logical categories in the palette:

Quick Reference - All Blocks

Here's a complete list of all available blocks in RAPFlow:

Common Blocks

• inject - Manually trigger flows and inject data
• debug - Monitor and debug data flow
• complete - Mark workflow completion
• catch - Handle errors and exceptions
• status - Display node status information
• link in - Create virtual input connections
• link call - Call linked subflows
• link out - Create virtual output connections
• comment - Add documentation to your flows
• extract_archive - Extract files from archives
• python_custom_shell - Execute Python scripts
• python_env_setup - Set up Python environment
• Text Processor - Process and manipulate text
• utils - Utility functions including bounding boxes

Function Blocks

• function - Execute custom JavaScript code
• switch - Route messages based on conditions
• change - Modify message properties
• range - Map values to different ranges
• template - Generate text from templates
• delay - Add delays to flow execution
• trigger - Advanced triggering and timing control
• filter - Filter messages based on criteria
• counter loop - Execute loops with counters
• array loop - Iterate through arrays
• while loop - Execute conditional loops

Network Blocks

• http in - Receive HTTP requests
• http response - Send HTTP responses
• http request - Make HTTP requests

Sequence Blocks

• split - Split strings or arrays
• join - Join arrays or strings
• sort - Sort arrays or objects
• batch - Group messages into batches

Parser Blocks

• csv - Work with CSV data
• html - Parse and manipulate HTML
• json - Parse and generate JSON
• xml - Work with XML data
• yaml - Parse and generate YAML

Storage Blocks

• write file - Write data to files
• read file - Read data from files
• watch - Monitor filesystem changes

Analysis Blocks

AI & ML Blocks

• llm guard - Security and safety for LLM interactions
• create conversation - Create conversation contexts
• fetch history - Retrieve conversation history
• LLM Query - Query large language models
• LLM Judge - Use LLM for evaluation and scoring
• VLM Query - Query vision language models

Document Processing Blocks

• Pre Process Document 2 - Prepare documents for processing
• Retrieve Matching Chunks - Find relevant document sections
• Document Classifier - Classify document types
• Document Question Answering - Extract answers from documents
• OCR - Optical character recognition
• PDF Processor - Process PDF documents
• Table Structure Recognition - Detect and extract table structures
• Template Matcher - Match documents to templates
• Signature Matcher - Detect and verify signatures
• Tiff Utils - Work with TIFF images

Classification & Extraction Blocks

• NLP Classifier - Classify text using natural language processing
• Entity Extractor - Extract named entities from text
• Image Classifier - Classify images
• Image Processor - Process and manipulate images
• Object Detector - Detect objects in images

Visualization & Output Blocks

• viewer - Display data and results
• image - Display and manipulate images

Workflow & Agent Blocks

• agent workgroup - Coordinate multiple AI agents

Verification & Storage Blocks

• Verifyme - Verification and validation services
• Write Data - Store data to various destinations

Getting Started with Blocks

1. Explore by Category: Browse blocks by their functional categories
2. Read Block Documentation: Each block has detailed configuration and usage information
3. Try Examples: Start with simple examples before building complex flows
4. Combine Blocks: Most workflows combine multiple blocks to achieve desired outcomes

Need Help?

• Check the individual block documentation for detailed configuration options
• See Use Cases for real-world examples
• Visit the FAQ for common questions and troubleshooting