A Claude Code plugin that detects when Claude edits or deletes files outside your request. Clear drift pauses on a question dialog; milder drift is flagged above the prompt with pull-back, fine, and allow-folder keys.

Review tools for permission checks and sensitive data handling.
54 projectsA Claude Code plugin that detects when Claude edits or deletes files outside your request. Clear drift pauses on a question dialog; milder drift is flagged above the prompt with pull-back, fine, and allow-folder keys.

Anthropic introduces Mods for Claude Code: packaged as TypeScript modules within plugins, they can hook into events such as tool calls, prompts, turns, slash commands, and UI rendering to rewrite or intercept them; installed via /plugin, requires version 2.1.287+. Official examples include
A Claude Code plugin that pairs Chrome browser profiles with cswap accounts and blocks Chrome tool actions when the active account does not match the paired profile.

BazaarLink is an OpenAI-compatible AI API gateway: one key for 140+ AI models, pay as you go, with budget limits.
bazaarlink.aiVisit sponsor ↗keeps secrets out of every transcript: api keys, tokens and private keys read as a stable placeholder in every kept row, and come back only inside the tool call that names them
Track what your AI coding tools cost
syrtis.nyanako.comVisit sponsor ↗A Claude Code mod that previews the blast radius of destructive Bash commands when they ask for permission: rm targets measured (files, size, what git can't restore) plus user-defined regex -> dry-run rules, shown as a line under the dialog with details in a side pane. Display only: it never changes the permission decision.

A Claude Code TypeScript mod that intercepts tool calls (Bash, Read) to block, escalate or rewrite them before execution, demonstrated within a broader agent safety pipeline POC.

A Claude Code function-hook plugin that blocks a `git add` from staging your whole `$HOME` work-tree (`dot add -A`, `dot add .`, `dot add *`). Ships with install instructions via `CLAUDE_CODE_PLUGIN_DIRS` or `--plugin-dir`, plus a configurable `dot-guard.wrappers` setting for custom dotfiles aliases.
A Claude Code mod that opens a side pane explaining each permission request in plain English with WHAT, WHY and RISK before you approve.
A Reddit user shares five small Claude Code hook mods they built after repeated close calls: a destructive-command guard that pauses on git reset --hard or rm -rf and lists uncommitted files at risk, a context-usage weather-forecast indicator above the prompt, an API-key filter for pasted secrets, a push-to-main and Friday-afternoon block, and a task board Claude can read and append to. The author notes hooks fail open — a crashing hook is skipped, so they added a catch that converts any crash into a deny — and admits the guards are pattern-matching only and not bulletproof.
A Claude Code plugin that blocks destructive shell commands and edits to secrets before they run, using function hooks on Bash, Edit, Write and NotebookEdit tool calls.
Top 3 by votes, then GitHub stars
No. Votes and stars measure popularity. Review the source and permission scan before installing.