Practice what you've learned with interactive projects. Each one ties directly to a lesson on the site — build real tools while reinforcing AI concepts.
Long AI conversations drift. Agent memory leaks. Embedding models vary wildly in quality. This advanced project ties all three problems together into one monitoring dashboard — built entirely by prompting Claude or ChatGPT. You'll walk away with a real tool and a much deeper understanding of how AI systems actually manage knowledge over time.
Combine three powerful AI concepts into one practical project: an agent that reasons in a loop, pulls chunks from a document store with high retrieval quality, and calls live third-party APIs securely via MCP OAuth. You'll see exactly how agents, RAG, and secure tool connections fit together in a real system.
Combine context engineering, AI agent design, and smart chunking strategies into one powerful tool. This advanced project walks you through building a document chunking agent that doesn't just split text — it engineers context around every chunk so your AI can actually reason over long documents. A real, working system you'll be proud of.
This advanced project ties together RAG pipelines, MCP client architecture, and context window management into a single cohesive tool. You'll prompt Claude to build a document assistant that retrieves relevant chunks from a local knowledge base, routes them through an MCP client, and intelligently trims context so nothing important gets cut off. By the end, you'll have a working system that shows you exactly how retrieval, tool connections, and token budgets work together in production AI apps.
Combine semantic search, function calling, and prompt injection defenses into one cohesive tool. You'll use AI to build a working document search interface that understands meaning — not just keywords — while keeping malicious content safely quarantined. A hands-on capstone for anyone serious about building AI-powered apps that are actually production-ready.
Combine RAG pipelines, MCP transport layers, and sampling parameter controls into one powerful knowledge base tool. You'll prompt Claude or ChatGPT to build a working app that retrieves context, connects via stdio or HTTP, and lets you tune randomness in real time. Perfect for anyone who wants to see how these three advanced concepts click together.
Long AI coding sessions break down when context gets stale, prompts drift, and handoffs get messy. This advanced project teaches you to build a Session Context Manager that auto-generates onboarding prompts, monitors context shrinkage, and structures operator/user system prompts — combining everything you learned this week into one practical tool.
Combine three powerful AI skills into one cohesive project: teach ChatGPT to remember your project preferences, use AI to generate and explain regex patterns for log parsing, and automate the whole workflow into your GitHub repo using MCP. By the end, you'll have a working log monitor that gets smarter every time you use it.
Combine three powerful AI techniques into one practical tool: reverse-engineer prompts from outputs you admire, use rubber duck debugging to unstick yourself when you're confused, and checkpoint your best discoveries so you can experiment without fear. You'll end up with a personal prompt archaeology lab you'll actually use.
Combine five core AI skills into one powerful project: give Claude controlled access to a real codebase with MCP Roots, delegate analysis tasks using subagent mode, verify understanding with the Explain It Back technique, and surface everything through a ChatGPT assistant tuned with custom instructions. The result is a personal onboarding tool you can point at any unfamiliar project and actually understand.
Put this week's skills together in one practical project: an AI-powered tool that analyzes a code snippet, writes automated tests for it, flags unfamiliar terms into a living glossary, and uses MCP sampling to decide what level of explanation you actually need. A real workflow you'll use again and again.
Combine Claude Code's slash command shortcuts with ChatGPT's structured output mode to build a dashboard generator that scans any project folder and produces a clean, organized HTML overview. You'll wire together two AI tools in a real workflow — one to gather data fast, one to format it precisely. Perfect for developers who want to stop doing status updates by hand.
In this project, you'll combine MCP file system access, AI prompt review techniques, and role stacking to build a Prompt Improvement Engine. Claude reads your raw prompt files from your computer, critiques them through multiple expert lenses, and saves polished, ready-to-use versions — all automatically.
Learn how to build a Codebase Onboarding Kit Generator that uses Claude Code to analyze a project, ChatGPT Voice Mode to brainstorm onboarding questions, and AI writing tools to produce a polished README and onboarding guide — all in one workflow.
In this project, you'll combine web browsing plugins, specialist system prompts, structured feedback loops, and conversation version control to build a real AI-powered tech news digest. You'll walk away with a working tool and a repeatable development workflow you can apply to any future AI project.
Combine context switching, task chunking, and code reading skills by building an interactive project planner. You'll practice changing AI conversation direction, breaking down complex requirements, and understanding generated code structure.
Combine AI refactoring, plugin chains, and learning notebooks into one powerful tool. You'll create a system that analyzes messy code, suggests improvements, tracks changes, and builds your personal refactoring knowledge base.
Build a practical coding assistant that combines AI personas, conversation saving, and Claude's variable naming expertise. You'll create reusable prompts and workflows that make naming variables effortless.
Combine everything you've learned about AI coding templates, plugin installation, and tool selection to build a comprehensive development workflow system. You'll create an organized system that grows with your skills.
Build your own intelligent coding assistant that combines the power of multiple AI tools. You'll create a system that generates code, explains how it works, and then reviews and improves it automatically using connected AI prompts.
Create an interactive troubleshooter that combines error debugging, tool selection, and workflow optimization into one intelligent assistant. Perfect for practicing multi-model AI strategies and understanding when to use different AI tools together.
Combine prompt shortcuts, code generation templates, and conversation threading to build a configurator that creates custom AI coding assistant setups. Perfect for applying multiple AI workflow techniques in one practical project.
Combine AI code review automation, workflow hooks, and structured learning plans into one powerful dashboard. Track your progress, automate quality checks, and manage your AI development workflow all in one place.
Build a comprehensive data exploration tool that combines prompt variation testing with ChatGPT's data analysis capabilities. You'll create a system that tries multiple prompt strategies and uses Claude's extended thinking to validate findings.
This project teaches you to build a complete issue tracker by combining GitHub Copilot Workspace for code generation, Claude's MCP for database operations, and token-efficient prompting techniques. You'll create a real application while learning to optimize AI tool costs and workflows.
Create a practical MCP server tester that validates configurations and simulates request/response cycles. You'll apply prompt templates, multi-file management, and custom AI assistants to build a real testing tool.
Create a visual code comparison tool that highlights differences between code versions. You'll learn to configure AI tools for code analysis, implement safe code sharing practices, and build a server component for enhanced functionality.
Build a simulator that lets you design, preview, and test Custom GPT configurations. You'll combine prompt journaling, system instructions, and ChatGPT memory concepts into one practical tool.
Build a practical git commit message generator by combining Windsurf IDE, Claude's MCP integration, and organized AI workflows. You'll create a tool that analyzes code diffs and generates conventional commit messages automatically.
Build a comprehensive prompt testing interface that lets you run two prompts simultaneously, score their outputs, and maintain a history of results. Perfect for optimizing your AI workflows.
Learn how to use Claude or ChatGPT to build a secure API key manager from scratch. Step-by-step prompts teach you to guide AI tools into writing real, working code.
Use Claude or ChatGPT to build a simulated chat interface from scratch. Learn to create typing indicators, conversation history, and localStorage persistence.
Learn to prompt AI tools to build a real-time prompt analyzer that scores across 5 quality criteria with instant feedback and improvement suggestions.
Use AI to create an interactive terminal simulator that teaches Claude Code commands through 10 challenges with hints and progress tracking.
Prompt AI to build a visual tool for designing multi-turn conversations with System, User, and Assistant roles, plus JSON export.
Use AI to create a code review tool with a 12-point checklist, animated score gauge, and quality assessment for AI-generated code.
Learn to prompt AI to build a multi-step wizard that generates valid MCP server configurations with OS-specific instructions and JSON export.
Use AI to create a reusable prompt template system with {{variable}} placeholders, pre-built templates, and custom template creation.
Prompt AI to build an interactive quiz that recommends Cursor, Copilot, or Claude Code based on user preferences and coding style.
Use AI to create a searchable directory of 20+ AI plugins and MCP servers with filtering, detail views, and favorites.
Prompt AI to build a personal learning tracker with progress bars, streak counting, study time logging, and data visualization.