learn-claude-code: This repository provides a hands-on tutorial to learn how modern AI agents work by building one from scratch, emphasizing the core loop of model-tool interaction. It distills complex agent concepts into simple, iterative versions, from basic Bash agents to sophisticated skill-based systems.; codeTree: codetree is an MCP server that uses tree-sitter to provide structured code understanding to AI coding agents, reducing token usage by 25x. It offers 23 tools across 10 languages with startup around 1 second. No vector DB, embedding model, or configuration needed.
Learning to build modern AI coding agents from scratch
Reduce token consumption for AI coding agents when exploring codebases