---
tags:
  - llm
  - agent
  - dev
---

# Key Areas of Rasa Syntax:

According to Gemini, 4 juli 2025

- **YAML Configuration:**
    
    Rasa uses YAML files (e.g., `config.yml`, `domain.yml`, `endpoints.yml`) to define the core components of your chatbot, including the NLU pipeline, domain, and endpoint configurations. 
    
- **Training Data:**
    
    Training data, typically in Markdown format, is used to train the NLU model. It includes intents, entities, and examples of user input. 
    
- **Stories:**
    
    Stories, also in Markdown, define the dialogue flows of your chatbot, mapping user inputs to bot responses and actions. 
    
- **Actions:**
    
    Custom actions, written in Python, define the logic that your chatbot executes in response to user input or specific events. 
    
- **Rules:**
    
    Rules are a simplified way to define dialogue flows, often used for simple scenarios or to handle specific edge cases. 
    
- **Forms:**
    
    Forms are used to collect multiple pieces of information from the user in a structured way. 
    
- **Responses:**
    
    Responses define the text, buttons, or other elements that your chatbot sends back to the user. 
    
- **Flows:**
    
    Flows are used in Rasa Pro to define the overall structure of your conversation, including steps, conditions, and actions. 
    
- **Conditions:**
    
    Conditions are used to control the flow of conversation based on slot values, user input, or other factors. 
    
- **Command Line Interface (CLI):**
    
    Rasa provides a CLI for various tasks, including training models, running servers, and inspecting assistants. 
    
- **Session Management:**
    
    Rasa handles session management using session configuration, expiration times, and slot carryover.