From 1f1fb03e71fd24cd1f641697aee1876864dd7e07 Mon Sep 17 00:00:00 2001
From: binrogithub <xpwilu@163.com>
Date: Sun, 8 Feb 2026 17:59:59 +0300
Subject: [PATCH] Add sanitization helper

---
 hc_agent/core/sanitize.py | 64 +++++++++++++++++++++++++++++++++++++++
 1 file changed, 64 insertions(+)
 create mode 100644 hc_agent/core/sanitize.py

diff --git a/hc_agent/core/sanitize.py b/hc_agent/core/sanitize.py
new file mode 100644
index 0000000000..02646dedf8
--- /dev/null
+++ b/hc_agent/core/sanitize.py
@@ -0,0 +1,64 @@
+"""Utilities for redacting sensitive data."""
+
+from __future__ import annotations
+
+import hashlib
+from typing import Any
+
+SENSITIVE_KEY_FRAGMENTS = (
+    "token",
+    "secret",
+    "cert",
+    "kubeconfig",
+)
+
+LARGE_STRING_THRESHOLD = 128
+
+
+def _is_sensitive_key(key: Any) -> bool:
+    if not isinstance(key, str):
+        return False
+    lowered = key.lower()
+    return any(fragment in lowered for fragment in SENSITIVE_KEY_FRAGMENTS)
+
+
+def _redacted_value(value: Any) -> str:
+    if isinstance(value, bytes):
+        digest = hashlib.sha256(value).hexdigest()
+        length = len(value)
+        return f"<redacted sha256={digest} len={length}>"
+    if isinstance(value, str):
+        digest = hashlib.sha256(value.encode("utf-8", "replace")).hexdigest()
+        length = len(value)
+        return f"<redacted sha256={digest} len={length}>"
+    return "<redacted>"
+
+
+def sanitize_obj(obj: Any) -> Any:
+    """Return a sanitized copy of nested data structures.
+
+    Redacts sensitive keys and large strings while preserving structure.
+    """
+    if isinstance(obj, dict):
+        sanitized: dict[Any, Any] = {}
+        for key, value in obj.items():
+            if _is_sensitive_key(key):
+                sanitized[key] = _redacted_value(value)
+            else:
+                sanitized[key] = sanitize_obj(value)
+        return sanitized
+    if isinstance(obj, list):
+        return [sanitize_obj(item) for item in obj]
+    if isinstance(obj, tuple):
+        return tuple(sanitize_obj(item) for item in obj)
+    if isinstance(obj, set):
+        return {sanitize_obj(item) for item in obj}
+    if isinstance(obj, bytes):
+        if len(obj) > LARGE_STRING_THRESHOLD:
+            return _redacted_value(obj)
+        return obj
+    if isinstance(obj, str):
+        if len(obj) > LARGE_STRING_THRESHOLD:
+            return _redacted_value(obj)
+        return obj
+    return obj