Skip to main content

Specification Overview

📘 v1.1 (Current)

PromptPack is a portable specification for packaging conversational AI systems into reusable, testable bundles. Think of it as a "container format" for AI applications—similar to how Docker containers package software, PromptPacks package everything needed to run sophisticated conversational AI.

Version Information

This documentation covers v1.1 of the PromptPack specification, which adds multimodal support. Looking for the previous version? View v1.0 docs →

Why PromptPacks?​

The Challenge​

Building production-ready conversational AI involves more than just writing prompts. You need:

  • Multiple specialized prompts for different scenarios (support, sales, technical help)
  • External tools that the AI can call (databases, APIs, calculators)
  • Shared resources like reusable text fragments and configurations
  • Safety guardrails to ensure appropriate responses
  • Version management to track changes and ensure compatibility
  • Testing frameworks to validate behavior across different models

Without a standard format, AI applications become fragmented, hard to maintain, and impossible to share reliably.

The Solution​

PromptPacks solve this by providing a single JSON file that contains everything needed to run a conversational AI system:

{
  "id": "customer-support",
  "name": "Customer Support Pack", 
  "version": "1.0.0",
  "prompts": {
    "support": { /* specialized for general support */ },
    "sales": { /* optimized for sales inquiries */ },
    "technical": { /* focused on technical issues */ }
  },
  "tools": { /* shared external functions */ },
  "fragments": { /* reusable text components */ }
}

Core Benefits​

🎯 Multi-Prompt Architecture​

Instead of one generic prompt trying to handle everything, PromptPacks let you create specialized prompts for specific tasks. A customer service pack might have separate prompts for billing questions, technical support, and sales inquiries—each optimized for its specific purpose while sharing common tools and configuration.

📦 Complete Packaging​

Everything needed to run your AI system is in one file. No more hunting for prompt templates, tool definitions, or configuration scattered across multiple files. Deploy once, run anywhere.

🔄 Reusability & Sharing​

PromptPacks are portable. Build a customer support pack once, then use it across different applications, teams, or even organizations. Share best practices through standardized, tested packages.

🛡️ Built-in Safety​

Each prompt can have its own validators (guardrails) to ensure safe, appropriate responses. Define content filters, length limits, and custom validation rules that travel with your prompts.

🧪 Testability​

PromptPacks include testing metadata—which models have been tested, success rates, performance metrics. Know before you deploy whether your pack works well with different AI providers.

⚡ Tool Integration​

Define external tools once, reference them from any prompt in the pack. Whether it's looking up customer data, performing calculations, or calling external APIs, tools are reusable across all prompts.

Real-World Use Cases​

Customer Service​

A complete customer service PromptPack might include:

  • Support prompt for general inquiries with ticket creation tools
  • Sales prompt for product questions with inventory lookup tools
  • Technical prompt for troubleshooting with diagnostic tools
  • Shared fragments for company policies and escalation procedures

Content Creation​

A content generation PromptPack could contain:

  • Blog writing prompt with SEO optimization tools
  • Social media prompt with character limit validators
  • Email marketing prompt with A/B testing tools
  • Brand voice fragments ensuring consistent tone across all content

Educational Assistant​

An educational PromptPack might feature:

  • Tutoring prompt with adaptive questioning techniques
  • Assessment prompt with grading rubrics and feedback tools
  • Research prompt with citation tools and fact-checking
  • Curriculum fragments aligned to learning standards

Design Philosophy​

PromptPacks follow key principles that make them powerful and practical:

Modularity: Each prompt handles one domain well rather than trying to do everything

Composability: Shared tools, fragments, and configuration reduce duplication

Portability: Works across different AI providers and runtime environments

Versioning: Track changes and maintain compatibility as your AI evolves

Observability: Built-in testing and performance tracking helps you optimize

Getting Started​

The PromptPack format is designed to be both human-readable and machine-executable. Whether you're hand-crafting prompts or generating them programmatically, the JSON structure provides the flexibility and power needed for production AI applications.

Ready to dive deeper? Explore the structure guide to understand how packs are organized, or jump to examples to see complete, real-world PromptPacks in action.