Claude Code Mods: Build Your Own Claude Code

Claude Code mods let you run your own JavaScript or TypeScript functions inside Claude Code. A mod can watch, change or stop a tool call before it runs, and it can draw its own panels in the terminal and the desktop app. In this video I explain what a Claude Code mod is and how it differs from hooks, skills and MCP servers. Then I show how to install mods from a marketplace and how I had Claude write a set of mods for my own workflow. The session-audit prompt I use (by Vox): https://x.com/Voxyz_ai/status/2105778540251644176 Claude Code mods docs: https://code.claude.com/docs/en/plugins/mods Getting started with Claude Code mods: https://claude.dev/blog/getting-started-with-claude-code-mods Mods walkthrough by Lydia Hallie (Claude Code team): https://x.com/lydiahallie/status/2106127556491821499 Claude Code changelog: https://github.com/anthropics/claude-code/blob/main/CHANGELOG.md DeepSeek Harness: https://github.com/deepseek-ai/deepseek-harness Mods run on your machine with your permissions. Read a mod, or run claude plugin validate on it, before you install it. My voice to text App: whryte.com Website: https://engineerprompt.ai/ RAG Beyond Basics Course: https://prompt-s-site.thinkific.com/courses/rag Signup for Newsletter, localgpt: https://tally.so/r/3y9bb0 Chapters: 0:00 Claude Code mods: Anthropic opens up Claude Code 1:39 What is a Claude Code mod and how does it work? 3:54 Mods vs hooks, skills and MCP: why it's a big deal 5:18 Your first mod + hot reload 6:32 Install a mod + a guard that stops rm -rf 7:42 Let Claude write your mods for you 9:15 Mod security and when to use a mod