claude-code-security-starter

English | 繁體中文 | 简体中文

A security starter pack for Claude Code beginners — CLAUDE.md rules + pre-built hooks that prevent credential leaks before they happen.

Why

Claude Code sends file contents to Anthropic servers whenever it reads a file. By default there are no guardrails — nothing stops it from reading your .env, .clasprc.json, or any file with "secret" in the name if you (or Claude) accidentally ask it to.

This pack closes those gaps with:

1. CLAUDE.md rules — instructions that tell Claude what not to touch, including Bash-based workarounds (because cat .env leaks just as much as Read .env)
2. block_sensitive_read hook — a hard OS-level block on the Read tool before Claude can even try
3. pre_push_safety hook — scans every git push for 25+ API key patterns before anything leaves your machine

Quick Start

This repo ships with a CLAUDE.md that tells Claude Code how to install everything. The easiest way:

git clone https://github.com/<YOUR_USERNAME>/claude-code-security-starter
cd claude-code-security-starter
claude  # open Claude Code in this directory

Then just say: "幫我安裝" or "install this" — Claude will copy the hooks, configure your forbidden directories, and merge the settings for you.

Manual install (without AI)

# 1. Copy hooks
mkdir -p ~/.claude/hooks
cp hooks/block_sensitive_read.sh ~/.claude/hooks/
cp hooks/pre_push_safety.sh ~/.claude/hooks/
chmod +x ~/.claude/hooks/*.sh

# 2. Edit your forbidden directories
nano ~/.claude/hooks/block_sensitive_read.sh   # fill in FORBIDDEN_DIRS

# 3. Merge settings.example.json into ~/.claude/settings.json

# 4. Copy and adapt the CLAUDE.md template to your project
cp CLAUDE.md.template /path/to/your/project/CLAUDE.md
# Replace all <PLACEHOLDER> values

Files

File	Purpose
CLAUDE.md.template	Security rules to paste into your project's CLAUDE.md
hooks/block_sensitive_read.sh	Blocks the Read tool on .env, credentials, and your custom forbidden paths
hooks/pre_push_safety.sh	Scans git push for 25+ API key / token regex patterns
settings.example.json	Claude Code hook wiring (merge into settings.json)

What Gets Blocked

block_sensitive_read.sh

• Any path you add to FORBIDDEN_DIRS (e.g. your ~/secrets/ folder)
• Any filename you add to FORBIDDEN_FILES
• .env and .env.*
• .clasprc* (Google clasp OAuth tokens)
• credentials.json, service-account.json
• Any filename containing secret, credential, token, or key

pre_push_safety.sh

Regex patterns for: AWS, Google API, Anthropic, OpenAI, GitHub, GitLab, Stripe, SendGrid, Twilio, Slack, HuggingFace, Perplexity, Replicate, npm, Telegram Bot, HashiCorp Vault, PEM private keys, database connection strings (MongoDB, PostgreSQL, MySQL), and generic api_key = "..." patterns.

How Hooks Work

Claude Code hooks are shell scripts that run at defined lifecycle events. A hook that exits with code 2 blocks the tool call and feeds the output back to Claude as context.

PreToolUse → hook runs → exit 2 → tool call cancelled
                       → exit 0 → tool call proceeds normally

More on hooks: https://code.claude.com/docs/hooks

License

MIT