Build an AI Research Agent in Claude Code (Live Demo) | Claude Code Tutorial

Build a powerful AI content research agent using Claude Code - no traditional coding required. In this live workshop, Dheeraj Sharma walks through the complete setup: MCP server configuration, context file architecture, and a working research agent that uses Perplexity AI, Firecrawl, and Gina AI to produce structured research briefs. 📚 What You'll Learn - How to set up Claude Code with MCP servers (Perplexity, Firecrawl, Gina AI) - The 3-layer architecture: Prerequisites → Context Files → Agent Instructions - How to build business context profiles (audience, voice, competitors, content pillars) - Structuring agent prompts with cost optimization rules - Running the research agent live and reviewing structured output (JSON + Markdown) - Tips for plan mode, cost control, and prompt structuring in Claude Code 🛠️ Tools & Services Used - Claude Code (Anthropic) - Terminal-based AI coding assistant - Perplexity AI - Real-time web research via MCP - Firecrawl - Web scraping and content extraction - Gina AI - Free backup scraping utility - Claude Sonnet model (recommended for research tasks) ⚙️ Prerequisites - Claude Code installed (Pro or Max subscription) - Perplexity API key - Firecrawl API key (500 free credits) - Basic familiarity with terminal/command line - A content business or niche you want to research 💡 Key Takeaways - You don't need to be a developer to use Claude Code for research automation - Context files (business profile, competitor intel, SEO focus) make your agent output 10x more relevant - Cost optimization rules in your agent prompt prevent burning through API credits - The research output includes competitor analysis, keyword opportunities, content gaps, and citations - Plan mode in Claude Code saves tokens and gives you control before execution 👥 About This Show This is Episode 1 of the One Shot Show - a weekly live stream by Wyndo (The AI Maker) and Dheeraj Sharma (GenAI Unplugged). Every Wednesday we build something practical with AI tools through live screen sharing. 0:00 Welcome & Show Introduction 5:15 Why Claude Code for Research 7:18 Screen Share & Prerequisites 10:23 MCP Server Setup 14:30 Why Perplexity & Firecrawl Matter 16:31 Architecture & Data Flow 20:38 Configuring the MCP JSON File 26:45 Building Context Files 33:56 Agent Prompt & Instructions 39:05 Cost Optimization Tips 43:16 Live Demo: Running the Agent 51:24 Reviewing Research Output 56:34 Q&A: Deep Research & Integrations 1:03:48 Wrap-Up & Next Steps 📥 Helpful Links: 🛠️ Check my AI Systems store → https://store.genaiunplugged.com/ ▶️ Watch the Full n8n Zero to Hero Course Playlist → https://tinyurl.com/3rw3x6hy 🌐 Subscribe to GenAI Unplugged Substack → https://genaiunplugged.substack.com/ 🤝 Let's connect on LinkedIn → https://www.linkedin.com/in/dheerajsharma14/ 📚 Browse all my courses → https://genaiunplugged.com/ 🙌 Like this video? Here's what to do next: ✅ Hit thumbs up to support the channel ✅ Subscribe for weekly AI automation tutorials & lessons ✅ Drop a comment: What AI system would you like me to build next? #ClaudeCode #AIAgent #ContentResearch