Build a Custom Native Event Sponsorship Scouting Subagent Inside Claude Code

I am looking for an experienced AI Engineer to build a custom native subagent called The Corporate Sponsorship & Event Scout Agent. We are a company looking to strategically sponsor existing public events, business expos, forums, and exhibitions happening in our country. Before we commit money to any sponsorship, this agent must act as our deep-dive intelligence tool to scout what events are happening, evaluate their full metrics, and determine if they are worth our investment. This entire system must be built exclusively inside Claude Code (Anthropic’s command-line agent tool) inside my local project directory. I do not want any separate coding languages (like Python/Node.js scripts) or external automation platforms. Everything must run natively through Claude Code's custom subagent framework. What the Event Scouting Agent Must Do: Live Country-Wide Web Scouting: The agent must use Claude Code’s Model Context Protocol (MCP) tool connections (like Tavily, Perplexity, or Firecrawl) to search local business directories, venue calendars, national expos, and upcoming event registries. Complete Event Metrics & Profile: For every event discovered, it must provide a comprehensive breakdown including estimated attendance, target demographics, cost/sponsorship tiers, and the exact attendee profile (who is coming, who we can meet on the floor, and what networking value exists). Full Strategic Analysis (Pros, Cons, Benefits, Losses): It must give clear advantages/benefits of sponsoring, potential disadvantages/losses (why an event might be a waste of money), and suggest creative activation ideas. Dual Sponsorship Categorization: It must automatically organize opportunities into: Corporate Branding: High-level summits, business forums, national conferences, or CSR initiatives. Product-Based Branding: Industry-specific events where we can target specific product lines directly. Full Verifiable Sources: The agent must provide full source links and URLs for every single event it flags so we can instantly verify the details. Strict Future Timeline: Focus entirely on upcoming active opportunities for the rest of 2026 and into 2027, completely ignoring past or expired events. Clean Markdown Tables: The final analysis must be delivered in beautifully structured Markdown tables so I can easily copy them into Excel or save them as a PDF. Your Deliverables: Custom Subagent File (event-scout.md): A custom agent definition file saved under .claude/agents/event-scout.md containing the full system persona, analysis guidelines, and strict output formatting rules. MCP Server Setup Configuration: A configured .mcp.json file linking a reliable search/scraping API so the subagent has live web-browsing capabilities to grab real-time data and links. Project Integration (CLAUDE.md): Updated project instructions so the main Claude Code terminal knows exactly when to call and route queries to the event-scout agent. Project Type: Fixed-Price Budget: $50 USD