
Block Dangerous Commands
Refuses rm -rf /, force-push to main, DROP TABLE, curl | sh and other destructive Bash commands before they run.
- toast
- Intercepts tool calls
- Changes the UI
tested with v2.1.291
$ claude --version ≥ 2.1.287
Tested mods for Claude Code with a demo, clear permission badges and a one-command install.

Refuses rm -rf /, force-push to main, DROP TABLE, curl | sh and other destructive Bash commands before they run.
tested with v2.1.291

Context-window fill, 5-hour and 7-day plan usage and session cost in the status line, plus a warning near the limit.
tested with v2.1.291

A "Done in 2m 14s" toast and a short chime when a long Claude turn finishes, plus an optional desktop notification.
tested with v2.1.291

Shows your git branch, changed-file count and ahead/behind status in a one-line band above the Claude Code prompt.
tested with v2.1.291

Counts Claude's tool calls per turn and session, shows the count by the spinner, and adds /tally for a breakdown.
tested with v2.1.291

Refuses rm -rf /, force-push to main, DROP TABLE, curl | sh and other destructive Bash commands before they run.
tested with v2.1.291

Context-window fill, 5-hour and 7-day plan usage and session cost in the status line, plus a warning near the limit.
tested with v2.1.291

Shows your git branch, changed-file count and ahead/behind status in a one-line band above the Claude Code prompt.
tested with v2.1.291
Claude Code mods are plugins that change how Claude Code looks and behaves while you work. A mod is a hooks module inside an ordinary Claude Code plugin: it listens to events such as tool.call, turn.complete or ui.render, and can draw a line above the prompt, write to the status line, show a toast, add a slash command or refuse a dangerous shell command before it runs. Mods arrived in Claude Code 2.1.287 and run in the terminal and the desktop app's Code tab.
ClaudeMods is an unofficial directory of mods we install and test ourselves. Before a page goes live, the mod passes claude plugin validate, its tests pass with claude plugin test, and we record it running in a real session on a current Claude Code build (2.1.291 at the moment). Each page shows that recording, the version and date of the test, and an install block generated from the same marketplace manifest Claude Code reads.
Each page also carries permission badges. They come from the events a mod handles and the API calls it makes, as reported by validation, not from the author's description. If a mod runs commands, reads files, uses the network or calls the model, you see it before you install. The "Permissions & safety" section explains why the mod needs each one.
Mods are plugins, so you manage them from the /plugin menu you already know: disable one, change its options in /config, or uninstall it. claude --safe-mode starts a session with every mod off. The site needs no account.
New to mods? Read What are Claude Code mods? and then How to install Claude Code mods. Good first installs come from the status line and safety categories.
Built a mod others would use? Submit it and we will test it, record a demo and list it with credit to you.