AI Integrations

AI integrations help assistants and agents move beyond conversation. They connect models to tools, APIs, files, databases, source control, project systems, and other real application context.

Start Here

MCP - Model Context Protocol A standard way for AI hosts to connect to tools, files, databases, APIs, and external systems.

API Tokens Provider keys for OpenAI, Anthropic, Gemini, Cohere, Hugging Face, Mistral, and related platforms.

LangChain Application orchestration with models, prompts, memory, retrievers, tools, and RAG workflows.

LangGraph Stateful agent workflows with branching, loops, supervisors, and more controlled multi-step execution.

CrewAI Role-based multi-agent collaboration patterns for teams of specialized AI agents.

Area	Use It For
MCP - Model Context Protocol	Standard tool and context connections for AI hosts such as Claude, Cursor, and Kiro.
API Tokens	Keys for OpenAI, Anthropic, Gemini, Cohere, Hugging Face, Mistral, and related platforms.
LangChain	Application-level orchestration with models, tools, prompts, memory, and retrievers.
LangGraph	Agent workflows where state, branching, and multi-step execution need more control.
CrewAI	Multi-agent collaboration patterns.

Integration Patterns

Tool access for actions such as creating issues, reading files, or querying systems.
Data access for documents, database schemas, tickets, and repository context.
API access for building custom applications on top of model providers.
Workflow orchestration for multi-step agents and retrieval-augmented generation.

Start with MCP when the goal is to connect an AI assistant to external tools. Start with API Tokens when the goal is to build or run your own AI-powered application.