Different Ways to Create Recipes

Explore all methods for creating PraisonAI recipes

How to Create Recipes Using CLI

1. Initialize New Recipe

   praisonai recipe init my-recipe
   
2. Navigate to Recipe

   cd my-recipe
   
3. Customize Generated Files

   Edit agents.yaml to match your requirements.

How to Create Recipes Manually

1. Create Directory Structure

   mkdir my-recipe
   cd my-recipe
   touch agents.yaml
   
2. Write agents.yaml

   framework: praisonai
   topic: "{{task}}"
   
   roles:
     agent:
       role: Assistant
       goal: Complete the task
       tasks:
         main:
           description: "{{task}}"
           expected_output: "Task result"
   
3. Run Recipe

   praisonai recipe run ./my-recipe --var task="Your task"
   

How to Create Recipes from Existing Agents

1. Define Agent in Python

   from praisonaiagents import Agent
   
   agent = Agent(
       name="researcher",
       role="Research Specialist",
       goal="Research topics thoroughly",
       tools=["internet_search"]
   )
   
2. Convert to YAML

   Create agents.yaml based on your agent:

   framework: praisonai
   topic: "{{task}}"
   
   roles:
     researcher:
       role: Research Specialist
       goal: Research topics thoroughly
       tools:
         - internet_search
       tasks:
         research:
           description: "Research {{task}}"
   

How to Create Recipes from GitHub

1. Fork Repository

   Fork the Agent-Recipes repository on GitHub.

2. Clone Your Fork

   git clone https://github.com/YOUR_USERNAME/Agent-Recipes
   cd Agent-Recipes
   
3. Create New Recipe Directory

   mkdir my-new-recipe
   # Create agents.yaml
   
4. Customize and Push

   # Edit files
   git add .
   git commit -m "Add my-new-recipe"
   git push
   

Recipe Creation Methods Comparison

Method	Best For	Complexity
CLI init	Quick start	Low
Manual	Full control	Medium
From Agent	Converting existing code	Medium
GitHub Fork	Contributing to community	Medium