AnthropicfeatureClaude Code Mods
Small TypeScript add-ons that change how Claude Code behaves and looks, for Claude Code users in the terminal or desktop app.
Details
- Mods can rewrite a prompt before it reaches the model, block or retry tool calls, approve or deny permission requests, hide secrets in tool output, and add or replace parts of the interface
- You write one in a few lines of TypeScript or ask Claude Code to build it. Mods ship inside plugins and install with /plugin
- Anthropic's examples include Token Weather (shows how full your context window is), Blast Radius (catches risky commands like rm -rf before they run), and Replay Theater (step through every file edit with /replay)
- The /diff view now runs as a built-in mod you can turn off or customize, and a built-in 'You should know' mod adds a side agent that flags things you or Claude might miss
- On Team and Enterprise plans a security mod loads first to block risky changes
- Mods run with full access to your machine and no sandbox, so only install them from sources you trust
Best for
Adding team guardrails or custom panels to Claude Code without waiting for Anthropic to build them
When to use it
You keep wishing Claude Code warned you about, showed you, or blocked something it doesn't today
Who benefits
Teams that standardize how they use Claude Code, and whoever sets up their tooling
Staying informed is great. Staying ahead is better.
Pick a topic and leave with something you created.
Get new module drops and weekly AI strategy.