Integrating with AI Agents
This section provides practical how-to guides for using Datafye's MCP (Model Context Protocol) servers to enable AI-assisted development workflows.
Overview
Datafye provides two complementary MCP servers that connect AI agents to different aspects of your development workflow:
Environment MCP Server - Connects to your deployed environment
SDK MCP Server - Provides SDK knowledge for algo development
These guides show you how to set up and use each MCP server for specific tasks.
Guides
Learn how to set up and use the Environment MCP Server to:
Explore your data cloud datasets and schemas conversationally
Query market data available in your deployment
Analyze backtest results and scorecards with AI assistance
Monitor deployed algo status and positions
Debug deployment issues interactively
Who this is for: Developers who have provisioned any Datafye deployment and want to explore their environment, analyze data, or debug issues using AI assistance.
Prerequisites:
Any Datafye deployment (features available based on what's deployed)
Any AI tool, agent, or LLM that integrates with MCP servers
Learn how to set up and use the SDK MCP Server to:
Learn the Datafye SDK conversationally
Generate SDK code through natural language descriptions
Get code examples and best practices on demand
Build strategies interactively with AI assistance
Understand SDK capabilities without constantly reading docs
Who this is for: Algo developers who want to write trading strategies using AI assistance, regardless of deployment type.
Prerequisites:
None! Works with any deployment type or even without a deployment
Any AI tool, agent, or LLM that integrates with MCP servers (most useful with coding assistants)
Node.js (to run the SDK MCP Server locally)
Using Both Together
For the most powerful AI-assisted workflow, use both MCP servers simultaneously:
Write code using the SDK MCP Server
Deploy and test your algo
Analyze results using the Environment MCP Server
Iterate based on insights
See each guide for instructions on configuring multiple MCP servers in your AI tool.
What is MCP?
Model Context Protocol (MCP) is an open standard developed by Anthropic that allows AI tools and agents to securely connect to external data sources and tools. Datafye's MCP servers implement this protocol to provide AI tools with deep context about your environment and development workflow.
For conceptual understanding, see:
The Datafye MCP Servers - Overview of both MCP servers
Environment MCP Server - Deployment-connected server
SDK MCP Server - SDK knowledge server
Supported AI Tools
These guides work with any AI tool, agent, or LLM that integrates with MCP servers, including:
Development IDEs - VS Code, Cursor, IntelliJ IDEA, Replit (with MCP-compatible extensions)
Standalone AI Tools - Claude (desktop/web with MCP), ChatGPT (if MCP-enabled), Claude Code
Other AI Agents - Any LLM or AI tool that supports the MCP protocol
Each guide provides configuration examples for common tools.
Next Steps
Choose the guide that matches your current task:
Exploring Deployments with AI - If you have a Datafye deployment and want to explore it with AI
Vibe Coding Algos with AI - If you want to learn and write SDK code with AI assistance
Last updated: 2025-10-23
Last updated