coldbox-ai-integration

Integrate AI chat, streaming, RAG, agents, and tool calling into ColdBox and BoxLang applications.

1|Updated Apr 10, 2026
One-click install
npx skills add https://github.com/ColdBox/skills --skill coldbox-ai-integration
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: coldbox-ai-integration
Source: https://github.com/ColdBox/skills/tree/main/coldbox/ai-integration
Command: npx skills add https://github.com/ColdBox/skills --skill coldbox-ai-integration

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Integrating modern AI capabilities into a ColdBox/BoxLang application involves many moving parts including provider configuration, streaming, memory and vector stores, document loaders, tool calling, and embedding pipelines; this Skill consolidates patterns and best practices to make those integrations practical and maintainable.

Core Features & Use Cases

  • Chat and Streaming: One-shot chat, async futures, and real-time streaming callbacks for server-sent events or live UIs.
  • Pipelines, Agents, and Tool Calling: Reusable pipelines, autonomous agents with memory and tools, and function-calling integrations for safe real-time data access.
  • RAG and Embeddings: Document loaders, chunking, embeddings, and vector memory stores for semantic search and context-aware responses.
  • Integration Patterns: Examples for injecting the AI service into handlers or services, using session or windowed memory for multi-tenant scenarios, and recommendations for production vector stores.

Quick Start

Install bx-ai via CommandBox and call aiChat from a ColdBox handler to summarize user content and store embeddings in a persistent vector memory.

Frequently Asked Questions about coldbox-ai-integration

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I integrate AI chat and streaming responses into a ColdBox application?

Integrate AI chat into ColdBox by injecting the AI service into handlers to enable one-shot chat, asynchronous futures, and real-time streaming callbacks for server-sent events. This allows live UI updates directly from your application logic.

What is the best way to add retrieval-augmented generation (RAG) and semantic search in BoxLang?

RAG in BoxLang uses document loaders, chunking, embeddings, and vector memory stores to achieve semantic document search and context-aware responses. You generate embeddings from your documents and store them in a persistent vector memory for retrieval.

Can I build autonomous agents with tool calling capabilities in ColdBox?

Yes, ColdBox supports building autonomous agents equipped with memory and tools. It provides function-calling integrations and tool or method bindings that allow agents to safely access real-time data and execute reusable pipelines.

Does ColdBox AI integration support different memory types for multi-tenant scenarios?

ColdBox supports windowed, session, and vector store memory types for multi-tenant scenarios. You can configure session or windowed memory to maintain isolated conversational context per user within your application handlers.

How do I manage provider configuration and model selection for AI in BoxLang?

Provider configuration and model selection in BoxLang are handled by satisfying specific provider settings within the AI service. This setup allows you to specify which AI models to use for generating chat, embeddings, or autonomous agent tasks.