Foot Evo Agent

The vision

Football clubs spend millions on scouting networks, data analysts, and recruitment teams. Foot Evo Agent brings that same level of analysis to anyone who cares about the game. It is a virtual football scout and team manager, built as an AI agent that powers intelligent features inside the FUT Evolution iOS app and can reason about players, squads, tactics, and transfer markets the way a seasoned sporting director would.

Agent architecture

The system is built on top of Claude Agent SDK with a multi-agent design. Rather than one monolithic prompt trying to do everything, the work is split across specialized sub-agents that each excel at a specific task.

The Scout

This sub-agent connects to player databases and league statistics through MCP servers. It can search for players by position, age, playing style, or statistical profile. When asked to find a replacement for an injured midfielder, the Scout queries multiple data sources, cross-references performance metrics, and returns a shortlist with detailed reasoning for each recommendation.

The Analyst

The Analyst sub-agent focuses on squad composition. It evaluates the current roster for weaknesses, identifies positions that lack depth, and flags age-profile risks. It understands concepts like squad balance, where you need a mix of experienced leaders and young talent, and positional versatility, where players who can cover multiple roles add more value than specialists.

The Tactician

This sub-agent builds optimal lineups based on player chemistry, recent form, and opponent analysis. It can simulate match outcomes using historical data and suggest formation changes that exploit specific weaknesses in the opposition. The Tactician also handles budget constraints, recommending transfer targets that deliver the best value within a set spending limit.

MCP server integration

Each data source is exposed through its own MCP server. Player databases, league standings, transfer market valuations, and match statistics all run as independent services that the agent can query on demand. This modular approach means adding a new data source is as simple as spinning up another MCP server, with no changes needed in the agent logic itself.

Why this matters

Foot Evo Agent is more than a football tool. It is a showcase of what AI agent orchestration can achieve when integrated into mobile apps with real users. The agent runs server-side but delivers results directly into the FUT Evolution iOS interface, giving users AI-powered scouting recommendations, squad suggestions, and tactical analysis right on their phone. The patterns used here, including sub-agent delegation, MCP-based tool use, and multi-step reasoning with budget constraints, transfer directly to enterprise use cases in hiring, supply chain, and portfolio management.

Technical highlights

• Claude Agent SDK for orchestrating multi-step reasoning across sub-agents
• MCP servers for modular, pluggable access to external data sources
• Sub-agent specialization: scouting, analysis, and tactical simulation each run independently
• Budget-aware recommendations that optimize squad improvement per dollar spent
• Python backend with async execution for parallel data fetching across multiple MCP servers
• Mobile integration layer that delivers agent results into the FUT Evolution iOS app in real time