DataOps-platform/deployment/scripts/scan_release_secrets.py

#!/usr/bin/env python3
"""Fail closed when a release tree contains secret or binary payloads."""

from __future__ import annotations

import re
import sys
from pathlib import Path

_FORBIDDEN_SUFFIXES = frozenset(
    {".der", ".jks", ".key", ".keystore", ".p12", ".pfx", ".pkcs12", ".pyc", ".pyo"}
)
_CONFIG_SUFFIXES = frozenset({".conf", ".env", ".ini", ".json", ".toml", ".txt", ".yaml", ".yml"})
_PRIVATE_BLOCK = re.compile(
    "-----BEGIN " + r"(?:ENCRYPTED |RSA |EC |OPENSSH )?PRIVATE KEY-----",
    re.IGNORECASE,
)
_API_SECRET = re.compile(r"\bsk-[A-Za-z0-9_-]{20,}\b")
_ASSIGNMENT = re.compile(
    r"(?im)^\s*(?:[A-Za-z0-9_]*_)?"
    r"(?:credential|password|passwd|secret|token|api_key|private_key)"
    r"\s*[:=]\s*['\"]?([^'\"\s#;,]{8,})"
)
_PLACEHOLDERS = ("${", "<", "change", "example", "placeholder", "replace-", "tbd")


def _reject(message: str) -> int:
    print(f"release secret scan rejected: {message}", file=sys.stderr)
    return 1


def scan(root: Path) -> int:
    if not root.is_absolute() or not root.is_dir() or root.is_symlink():
        return _reject("root must be an existing absolute directory")
    for path in sorted(root.rglob("*")):
        relative = path.relative_to(root)
        if path.is_symlink():
            return _reject(f"symbolic link {relative}")
        if path.is_dir():
            if path.name == "__pycache__":
                return _reject(f"cache directory {relative}")
            continue
        if not path.is_file() or path.suffix.casefold() in _FORBIDDEN_SUFFIXES:
            return _reject(f"forbidden file type {relative}")
        try:
            raw = path.read_bytes()
            if b"\x00" in raw:
                return _reject(f"binary file {relative}")
            text = raw.decode("utf-8")
        except (OSError, UnicodeDecodeError):
            return _reject(f"unreadable or non-UTF-8 file {relative}")
        if _PRIVATE_BLOCK.search(text) or _API_SECRET.search(text):
            return _reject(f"private key or API secret shape in {relative}")
        if path.suffix.casefold() in _CONFIG_SUFFIXES or path.name.startswith(".env"):
            for match in _ASSIGNMENT.finditer(text):
                value = match.group(1).casefold()
                if value in {"false", "none", "null", "true"}:
                    continue
                if any(marker in value for marker in _PLACEHOLDERS):
                    continue
                return _reject(f"literal credential shape in {relative}")
    return 0


def main() -> int:
    if len(sys.argv) != 2:
        print("usage: scan_release_secrets.py ABSOLUTE_RELEASE_ROOT", file=sys.stderr)
        return 2
    return scan(Path(sys.argv[1]))


if __name__ == "__main__":
    raise SystemExit(main())