From 082ed8b1ae497c58879ea8bcb84b8ab3dd31486d Mon Sep 17 00:00:00 2001
From: dmytryk-here <6403375+dmytryk-here@users.noreply.github.com>
Date: Tue, 16 Jun 2026 08:52:10 +0800
Subject: [PATCH] Sanitize dirty JSON tool responses

---
 helpers/extract_tools.py               | 31 +++++++++++++++++++++++++-
 tests/test_extract_tools_json_parse.py | 23 +++++++++++++++++++
 2 files changed, 53 insertions(+), 1 deletion(-)
 create mode 100644 tests/test_extract_tools_json_parse.py

diff --git a/helpers/extract_tools.py b/helpers/extract_tools.py
index 39e8378553..3fd767c8d1 100644
--- a/helpers/extract_tools.py
+++ b/helpers/extract_tools.py
@@ -4,11 +4,40 @@
 from helpers.modules import load_classes_from_file, load_classes_from_folder # keep here for backwards compatibility
 from typing import Any
 
+
+def _sanitize_for_json_parsing(content: str) -> str:
+    """
+    Remove common LLM output artifacts that break JSON parsing:
+    markdown fences, XML-like wrappers, text outside JSON, and leading/trailing whitespace.
+    """
+    if not isinstance(content, str):
+        return ""
+
+    # 1. Strip markdown code blocks (```json ... ``` and ``` ... ```)
+    cleaned = re.sub(r'^```[a-zA-Z]*\s*\n?', '', content, flags=re.MULTILINE)
+    cleaned = re.sub(r'\n?```\s*$', '', cleaned, flags=re.MULTILINE)
+
+    # 2. Strip XML-style wrappers (<invoke>...</invoke>)
+    cleaned = re.sub(r'<\s*invoke\b[^>]*>.*?</\s*invoke\s*>', '', cleaned, flags=re.DOTALL)
+    cleaned = re.sub(r'<\s*/\s*invoke\s*>', '', cleaned, flags=re.DOTALL)
+
+    # 3. Remove function-style wrappers (functions.tool_name:123)
+    cleaned = re.sub(r'functions\.\w+\s*:\s*\w+\s*', '', cleaned)
+
+    # 4. Remove leading text before the first '{' if any
+    first_brace = cleaned.find('{')
+    if first_brace > 0:
+        cleaned = cleaned[first_brace:]
+
+    return cleaned
+
+
 def json_parse_dirty(json: str) -> dict[str, Any] | None:
     if not json or not isinstance(json, str):
         return None
 
-    ext_json = extract_json_object_string(json.strip())
+    sanitized = _sanitize_for_json_parsing(json)
+    ext_json = extract_json_object_string(sanitized.strip())
     if ext_json:
         try:
             data = DirtyJson.parse_string(ext_json)
diff --git a/tests/test_extract_tools_json_parse.py b/tests/test_extract_tools_json_parse.py
new file mode 100644
index 0000000000..9456113706
--- /dev/null
+++ b/tests/test_extract_tools_json_parse.py
@@ -0,0 +1,23 @@
+from __future__ import annotations
+
+import sys
+from pathlib import Path
+
+PROJECT_ROOT = Path(__file__).resolve().parents[1]
+if str(PROJECT_ROOT) not in sys.path:
+    sys.path.insert(0, str(PROJECT_ROOT))
+
+from helpers.extract_tools import json_parse_dirty
+
+
+def test_json_parse_dirty_sanitizes_common_llm_wrappers() -> None:
+    payload = """```json
+<invoke>discard this wrapper artifact</invoke>
+functions.search:123
+{"tool_name":"search","tool_args":{"query":"agent zero"}}
+```"""
+
+    assert json_parse_dirty(payload) == {
+        "tool_name": "search",
+        "tool_args": {"query": "agent zero"},
+    }