Agent frameworks
Build AI agents that interact with Box content.Box AI Agents Toolkit
A Python library for building AI agents that interact with Box. Provides
authentication, content interaction, and more.
Pydantic AI
Build AI agents using the Pydantic AI framework with the Box MCP server.
RAG frameworks
Include Box content in retrieval-augmented generation pipelines. These loaders copy file text into your index. After that, Box no longer enforces who can open each file. Unless you filter results, a query can return content the current user cannot access in Box. See keep an exported index permission-aware.LangChain
Include Box content in your AI
workflows.
LangChain.js
Use the Box loader for LangChain.js in JavaScript-based AI workflows.
LlamaIndex
Ingest Box content into LlamaIndex retrieval pipelines.
Vector databases
Connect Box content to vector databases for semantic search and similarity matching.Pinecone
Connect Box content to Pinecone for vector search workflows.
Weaviate
Build RAG workflows with Box content and the Weaviate vector database.
Enterprise platforms
Connect Box to enterprise platforms for agent-based and data workflows.Box for Agentforce
Extend Box for Salesforce with reusable Agentforce actions that automate
workflows and enhance agent-based processes.
Snowflake Openflow
Integrate Box with Snowflake using the Openflow Connector.
Airbyte
Extract Box data using the Airbyte source connector for data pipelines.
Learning resources
Tutorials and walkthroughs for building AI-powered workflows with Box.Build AI Apps with MCP Servers
Online course from DeepLearning.AI on building AI applications with MCP
servers and Box files.
Multi-agent workflows with OpenAI
Build multi-agent workflows using the OpenAI SDK and Box.
LangChain MCP Adapters
Use the Box MCP server with LangChain MCP Adapters for agent
orchestration.
Document generation with Box MCP
Build AI-powered document generation with the Box MCP server and Pydantic
AI.