examples

Programmable Prompt Engine (PPE) Script Examples 🤖

This directory contains a collection of example scripts to demonstrate the capabilities of the Programmable Prompt Engine (PPE). Each example showcases a different use case and highlights how to effectively leverage PPE's features.

Example Scripts

• calculator:
  • A simple calculator agent that demonstrates how to use the -> operator to connect script outputs to subsequent scripts, creating complex workflows.
  • Workflow:
    1. The LLM receives a mathematical expression (content).
    2. It processes the expression and outputs a thought process (thinking).
    3. The extract-calc-result script then extracts the final calculated result from the LLM's output.
• extract-calc-result:
  • This script defines an AI workflow for extracting the result of a calculated math problem from text.
  • Input: The script receives the LLM's output (content) as input.
  • Output: A JSON object containing:
    • result: The final calculated result (number, object, boolean, or string).
    • steps: A list of steps taken by the LLM during the calculation process.
    • answer: The LLM's final answer.
• resolve-math-problem.ai:
  • A demo agent for resolving math problems.
  • Input: Takes a user's math problem as a string.
  • Workflow:
    • If content is provided, it starts the problem-solving process.
    • If no content is provided, it greets the user and introduces itself as a math assistant.
• call-translator:
  • This example showcases how to invoke the translator lib script.
• char-dobby:
  • An interactive demo featuring a character agent named Dobby.
  • Demonstrates the use of the type keyword in the configuration section.
• recipe:
  • A Multilingual Recipe Assistant demo
  • Key Features:
    • Multilingual Support: Understands and responds in multiple languages.
    • Cuisine Specificity: Provides recipes from a specific cuisine.
    • Ingredient-Based Recommendations: Suggests recipes based on available ingredients.
    • JSON Output Format: Optionally outputs responses in JSON format.
  • Input:
    • ingredients: A list of available ingredients.
    • cuisine: (Optional) Desired cuisine type.
    • lang: (Optional) User's preferred language.
    • json: (Optional) Boolean flag for JSON output.
  • Output:
    • An array of recipe objects, each containing:
      • recipeName: Name of the recipe.
      • instructions: Cooking steps.
      • ingredients: Ingredient list with name and amount.
      • reason: Rationale behind the suggestion.
• translator-simple:
  • A simple translator agent demo.
  • Translation Task: Translate text from one language to another.
  • Input:
    • content: Text to be translated (required).
    • target: Target language (required).
    • lang: Original language (optional).
  • Output:
    • A JSON object containing:
    • translation: The translated text.
    • original: The original text.
    • lang: The original language.
    • target_lang: The target language.
    • reason: Explanation of the translation process (optional).
  • Parameters:
    • continueOnLengthLimit: Allows translation to continue even if the text exceeds a length limit.
    • maxRetry: Specifies the maximum number of retries for the translation.
    • response_format: Sets the output format to JSON.