From 8ee4676a683748a878f95993829c3064d458f24b Mon Sep 17 00:00:00 2001
From: ahmifte <ahmediftekhar@g.ucla.edu>
Date: Wed, 10 Jun 2026 17:33:53 -0700
Subject: [PATCH] chore: add seal pre-commit hook for secret scanning

---
 .githooks/gitleaks.toml |  83 +++++++++++++++++++++++++++++
 .githooks/pre-commit    | 115 ++++++++++++++++++++++++++++++++++++++++
 2 files changed, 198 insertions(+)
 create mode 100644 .githooks/gitleaks.toml
 create mode 100755 .githooks/pre-commit

diff --git a/.githooks/gitleaks.toml b/.githooks/gitleaks.toml
new file mode 100644
index 0000000..75e4f95
--- /dev/null
+++ b/.githooks/gitleaks.toml
@@ -0,0 +1,83 @@
+title = "seal gitleaks config"
+
+[extend]
+useDefault = true
+
+# Extra patterns beyond gitleaks defaults — tuned for common dev mistakes.
+[[rules]]
+id = "generic-env-assignment"
+description = "Likely secret assigned in an env file (not .env.example)"
+regex = '''(?i)(api[_-]?key|secret|token|password|passwd|credential)\s*=\s*['"]?[a-zA-Z0-9_\-./]{8,}['"]?'''
+tags = ["key", "env"]
+
+[[rules]]
+id = "stripe-restricted-key"
+description = "Stripe restricted API key"
+regex = '''rk_(live|test)_[0-9a-zA-Z]{10,}'''
+tags = ["stripe", "key"]
+
+[[rules]]
+id = "stripe-secret-key"
+description = "Stripe secret API key"
+regex = '''sk_(live|test)_[0-9a-zA-Z]{10,}'''
+tags = ["stripe", "key"]
+
+[[rules]]
+id = "stripe-webhook-secret"
+description = "Stripe webhook signing secret"
+regex = '''whsec_[0-9a-zA-Z]{10,}'''
+tags = ["stripe", "key"]
+
+[[rules]]
+id = "openai-api-key"
+description = "OpenAI API key"
+regex = '''sk-[a-zA-Z0-9]{20,}'''
+tags = ["openai", "key"]
+
+[[rules]]
+id = "github-pat"
+description = "GitHub personal access token"
+regex = '''ghp_[0-9a-zA-Z]{20,}'''
+tags = ["github", "key"]
+
+[[rules]]
+id = "github-oauth"
+description = "GitHub OAuth token"
+regex = '''gho_[0-9a-zA-Z]{20,}'''
+tags = ["github", "key"]
+
+[[rules]]
+id = "github-fine-grained-pat"
+description = "GitHub fine-grained PAT"
+regex = '''github_pat_[0-9a-zA-Z_]{20,}'''
+tags = ["github", "key"]
+
+[[rules]]
+id = "aws-access-key"
+description = "AWS access key ID"
+regex = '''AKIA[0-9A-Z]{16}'''
+tags = ["aws", "key"]
+
+[[rules]]
+id = "slack-token"
+description = "Slack token"
+regex = '''xox[baprs]-[0-9A-Za-z\-]{10,}'''
+tags = ["slack", "key"]
+
+[allowlist]
+description = "Safe placeholders and example files"
+paths = [
+  '''\.env\.example$''',
+  '''\.env\.sample$''',
+  '''README\.md$''',
+  '''SECURITY\.md$''',
+  '''gitleaks\.toml$''',
+  '''hooks/pre-commit$''',
+]
+regexes = [
+  '''your-[a-z-]+-key''',
+  '''replace-with''',
+  '''example\.com''',
+  '''xxxxxxxx''',
+  '''<[A-Z_]+>''',
+]
diff --git a/.githooks/pre-commit b/.githooks/pre-commit
new file mode 100755
index 0000000..1e3112d
--- /dev/null
+++ b/.githooks/pre-commit
@@ -0,0 +1,115 @@
+#!/usr/bin/env bash
+# seal — pre-commit secret scanner
+# Blocks commits that stage secrets, private env files, or high-risk key patterns.
+set -euo pipefail
+
+ROOT="$(git rev-parse --show-toplevel)"
+HOOK_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+
+RED='\033[0;31m'
+YELLOW='\033[1;33m'
+NC='\033[0m'
+
+fail() {
+  echo -e "${RED}seal: commit blocked — $1${NC}" >&2
+  exit 1
+}
+
+warn() {
+  echo -e "${YELLOW}seal: warning — $1${NC}" >&2
+}
+
+# --- 1. Block staging of files that should never be committed ---
+BLOCKED_FILES=(
+  ".env"
+  ".env.local"
+  ".env.production"
+  ".env.development"
+  "credentials.json"
+  "service-account.json"
+  "id_rsa"
+  "id_ed25519"
+  "*.pem"
+  "*.p12"
+  "*.pfx"
+)
+
+STAGED=$(git diff --cached --name-only --diff-filter=ACMR)
+
+for file in $STAGED; do
+  base=$(basename "$file")
+  for blocked in "${BLOCKED_FILES[@]}"; do
+    if [[ "$blocked" == *"*"* ]]; then
+      ext="${blocked#\*}"
+      if [[ "$base" == *"$ext" ]]; then
+        fail "refusing to commit '$file' ($ext files must stay local)"
+      fi
+    elif [[ "$base" == "$blocked" ]]; then
+      fail "refusing to commit '$file' (add it to .gitignore; only .env.example belongs in git)"
+    fi
+  done
+done
+
+# --- 2. Prefer gitleaks when installed (best coverage) ---
+if command -v gitleaks >/dev/null 2>&1; then
+  CONFIG="$HOOK_DIR/gitleaks.toml"
+  if [[ ! -f "$CONFIG" ]]; then
+    CONFIG="$ROOT/gitleaks.toml"
+  fi
+  if [[ ! -f "$CONFIG" ]]; then
+    CONFIG=""
+  fi
+  if [[ -n "$CONFIG" ]]; then
+    gitleaks protect --staged --redact --verbose --config "$CONFIG" --source "$ROOT"
+  else
+    gitleaks protect --staged --redact --verbose --source "$ROOT"
+  fi
+  echo "seal: gitleaks scan passed"
+  exit 0
+fi
+
+# --- 3. Fallback: pattern scan on staged content (no gitleaks required) ---
+PATTERNS=(
+  'sk_live_[0-9a-zA-Z]{10,}'
+  'sk_test_[0-9a-zA-Z]{10,}'
+  'rk_live_[0-9a-zA-Z]{10,}'
+  'rk_test_[0-9a-zA-Z]{10,}'
+  'whsec_[0-9a-zA-Z]{10,}'
+  'sk-[a-zA-Z0-9]{20,}'
+  'ghp_[0-9a-zA-Z]{20,}'
+  'gho_[0-9a-zA-Z]{20,}'
+  'github_pat_[0-9a-zA-Z_]{20,}'
+  'AKIA[0-9A-Z]{16}'
+  'xox[baprs]-[0-9A-Za-z\-]{10,}'
+)
+
+FOUND=0
+for file in $STAGED; do
+  # Skip allowlisted paths
+  case "$file" in
+    *.env.example|*.env.sample|README.md|SECURITY.md|gitleaks.toml) continue ;;
+  esac
+
+  if [[ ! -f "$ROOT/$file" ]]; then
+    continue
+  fi
+
+  # Skip binary files
+  if file -b --mime-type "$ROOT/$file" 2>/dev/null | grep -qE '^(image|video|audio|application/(zip|gzip|pdf|octet-stream))'; then
+    continue
+  fi
+
+  for pattern in "${PATTERNS[@]}"; do
+    if grep -qE "$pattern" "$ROOT/$file" 2>/dev/null; then
+      echo -e "${RED}seal: possible secret in $file (matched: $pattern)${NC}" >&2
+      FOUND=1
+    fi
+  done
+done
+
+if [[ "$FOUND" -eq 1 ]]; then
+  fail "secret-like value detected in staged files. Install gitleaks for deeper scanning: brew install gitleaks"
+fi
+
+warn "gitleaks not installed — using basic pattern scan only. Run: brew install gitleaks"
+echo "seal: basic scan passed"