Claude 3 7 Sonnet Starter Pack

Claude 3.7 Sonnet Starter Pack 🚀

Look, Anthropic's Claude-3.7-Sonnet is a powerful, hybrid reasoning, CRASHOUT LLM.

Their new CLAUDE CODE is THE best showcase of an [effective AI AGENT](https://www.anthropic.com/research/building-effective-agents) TO DATE.

Let's breakdown their capabilities and see how you can use it in your engineering work and play.

*Check out [IndyDevDan YouTube Channel](https://youtu.be/jCVO57fZIfM) for the full breakdown of this starter pack and for more actionable insights on AI Agents, LLMs, and AI Coding.*


Overview

This starter pack provides a collection of simple, self-contained examples showcasing the capabilities of Claude 3.7 Sonnet, Anthropic's latest and most powerful model. Each script demonstrates a specific feature or capability, making it easy to understand and integrate into your own projects.

Installation - Install [uv](https://docs.astral.sh/uv/getting-started/installation/)

uv is THE modern package manager for Python.

macos + linux

brew install uv

windows

powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex"

Usage - Single File Sonnet Scripts

See each file to see usage instructions. Start with `simple_prompt.py` and work your way up.

• simple_prompt.py - Basic prompt to Claude 3.7 Sonnet
• simple_structured_output.py - Demonstrates structured JSON output
• simple_tool_use.py - Shows basic tool use integration
• prompt_with_extended_thinking.py - Implements extended thinking for complex reasoning
• prompt_with_extended_thinking_and_streaming.py - Combines extended thinking with streaming responses
• prompt_with_extended_thinking_tool_use.py - Integrates extended thinking with tool use
• prompt_with_extended_output_and_extended_thinking_and_streaming.py - Showcases extended output combined with extended thinking and streaming
• `agent_bash_an