anthropics/cwc-workshops

Anthropic's cwc-workshops repo is a collection of hands-on training materials from their "Code with Claude" workshops. It's a set of guided exercises that teach you how to build, refine, and deploy AI agents using Claude, Anthropic's AI model. Think of it as a workbook full of starter projects that walk you through building practical AI tools, step by step. The repo covers a wide range of real-world scenarios. For example, one workshop teaches you how to build a dashboard that helps a site reliability engineer investigate incidents by searching through massive log files to pinpoint the problematic code change. Another walks you through breaking a massive, unwieldy AI prompt into smaller, manageable pieces that work together. There's also a competitive exercise where you configure an AI agent to play a game and collect diamonds, balancing performance against the cost of running the AI. These materials are designed for developers, product managers, or technical founders who want to move beyond simple chatbot interactions and build autonomous AI systems. If you want an AI agent that can read financial filings and write a weekly research memo, or one that generates PowerPoint decks and grades its own work for quality, these workshops provide the blueprints. They emphasize a test-driven approach: instead of guessing whether a prompt change improved the AI, you score every iteration against a checklist to measure actual progress. A notable thread running through the workshops is the progression from simple to sophisticated. Several exercises start with a basic, bare-bones AI agent and then layer on advanced features like persistent memory (so the agent remembers things across sessions) or multi-agent coordination (where a "manager" AI delegates tasks to specialist AIs). The materials also cover practical concerns like choosing the right model settings to balance quality, speed, and cost. Note that this is an archived resource, the repo states it is not maintained and does not accept contributions.