Fix RCE vulnerability and harden security

- Replace eval() with ast.literal_eval() in pdf_utils.py to fix
  unauthenticated remote code execution via crafted PDF uploads
  (reported by OX Security)
- Sanitize HTML output with DOMPurify to prevent XSS
- Restrict CORS origins (configurable via CORS_ORIGINS env var)
- Suppress raw exception details in API error responses
- Cap Image.MAX_IMAGE_PIXELS to prevent decompression bomb DoS
- Add security regression test suite

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

.env.example

@@ -11,6 +11,9 @@ FRONTEND_PORT=3000
 MODEL_NAME=deepseek-ai/DeepSeek-OCR
 HF_HOME=/models
 
+# CORS Configuration (comma-separated origins, defaults to http://localhost:3000)
+CORS_ORIGINS=http://localhost:3000
+
 # Upload Configuration
 MAX_UPLOAD_SIZE_MB=100
 

backend/main.py

@@ -86,11 +86,14 @@ app = FastAPI(
 )
 
 # CORS middleware for React frontend
+CORS_ORIGINS = env_config("CORS_ORIGINS", default="").split(",")
+CORS_ORIGINS = [o.strip() for o in CORS_ORIGINS if o.strip()]
+
 app.add_middleware(
     CORSMiddleware,
-    allow_origins=["*"],
+    allow_origins=CORS_ORIGINS if CORS_ORIGINS else ["http://localhost:3000"],
     allow_credentials=True,
-    allow_methods=["*"],
+    allow_methods=["GET", "POST"],
     allow_headers=["*"],
 )
 
@@ -373,7 +376,8 @@ async def ocr_inference(
         })
     
     except Exception as e:
-        raise HTTPException(status_code=500, detail=f"{type(e).__name__}: {str(e)}")
+        print(f"OCR inference error: {type(e).__name__}: {str(e)}")
+        raise HTTPException(status_code=500, detail="An internal error occurred during OCR processing.")
     
     finally:
         if tmp_img:
@@ -573,9 +577,9 @@ async def process_pdf(
 
     except Exception as e:
         import traceback
-        print(f"❌ Error processing PDF: {e}")
+        print(f"Error processing PDF: {e}")
         print(traceback.format_exc())
-        raise HTTPException(status_code=500, detail=f"{type(e).__name__}: {str(e)}")
+        raise HTTPException(status_code=500, detail="An internal error occurred during PDF processing.")
 
 if __name__ == "__main__":
     host = env_config("API_HOST", default="0.0.0.0")

backend/pdf_utils.py

@@ -3,6 +3,7 @@ PDF Processing Utilities for DeepSeek OCR
 Handles PDF to image conversion and batch processing
 """
 
+import ast
 import io
 import re
 from typing import List, Tuple, Dict, Any
@@ -39,8 +40,8 @@ def pdf_to_images_high_quality(pdf_bytes: bytes, dpi: int = 144) -> List[Image.I
         # Render page to pixmap
         pixmap = page.get_pixmap(matrix=matrix, alpha=False)
 
-        # Allow large images
-        Image.MAX_IMAGE_PIXELS = None
+        # Allow reasonably large images (200 megapixels) but not decompression bombs
+        Image.MAX_IMAGE_PIXELS = 200_000_000
 
         # Convert to PIL Image
         img_data = pixmap.tobytes("png")
@@ -130,7 +131,7 @@ def parse_coordinates(ref_text: Tuple, image_width: int, image_height: int) -> D
     """
     try:
         label_type = ref_text[1]
-        cor_list = eval(ref_text[2])
+        cor_list = ast.literal_eval(ref_text[2])
 
         # Scale coordinates from 0-999 to actual pixels
         scaled_boxes = []

backend/test_security.py

@@ -0,0 +1,150 @@
+"""
+Security regression tests for the eval() RCE vulnerability (OX Security disclosure).
+
+The vulnerability allowed arbitrary code execution via crafted OCR output
+that was passed to eval() in parse_coordinates(). The fix uses ast.literal_eval()
+which only allows literal data structures.
+
+This test is self-contained and does not require backend dependencies.
+
+Run: python test_security.py
+"""
+
+import ast
+
+
+def parse_coordinates(ref_text, image_width, image_height):
+    """
+    Minimal reproduction of pdf_utils.parse_coordinates using the patched code.
+    This mirrors the fixed version that uses ast.literal_eval() instead of eval().
+    """
+    try:
+        label_type = ref_text[1]
+        cor_list = ast.literal_eval(ref_text[2])
+
+        scaled_boxes = []
+        for points in cor_list:
+            x1, y1, x2, y2 = points
+            scaled_box = [
+                int(x1 / 999 * image_width),
+                int(y1 / 999 * image_height),
+                int(x2 / 999 * image_width),
+                int(y2 / 999 * image_height)
+            ]
+            scaled_boxes.append(scaled_box)
+
+        return {
+            'label': label_type,
+            'boxes': scaled_boxes
+        }
+    except Exception as e:
+        print(f"  [Blocked] {type(e).__name__}: {e}")
+        return None
+
+
+def test_legitimate_coordinates():
+    """Verify that normal coordinate parsing still works."""
+    ref_text = ("full_match", "text", "[[312, 339, 480, 681]]")
+    result = parse_coordinates(ref_text, 1000, 1000)
+
+    assert result is not None, "Legitimate coordinates should parse successfully"
+    assert result['label'] == 'text'
+    assert len(result['boxes']) == 1
+    print("PASS: Legitimate coordinates parse correctly")
+
+
+def test_multiple_boxes():
+    """Verify multiple bounding boxes still work."""
+    ref_text = ("full_match", "image", "[[100, 200, 300, 400], [500, 600, 700, 800]]")
+    result = parse_coordinates(ref_text, 1000, 1000)
+
+    assert result is not None, "Multiple boxes should parse successfully"
+    assert len(result['boxes']) == 2
+    print("PASS: Multiple bounding boxes parse correctly")
+
+
+def test_rce_blocked_import_os():
+    """The original exploit: __import__('os').system('...') must be blocked."""
+    malicious = "__import__('os').system('echo HACKED')"
+    ref_text = ("full_match", "exploit", malicious)
+    result = parse_coordinates(ref_text, 1000, 1000)
+
+    assert result is None, "Code execution payload should be rejected"
+    print("PASS: __import__('os').system() payload is blocked")
+
+
+def test_rce_blocked_exec():
+    """exec() based payloads must be blocked."""
+    malicious = "exec('import os; os.system(\"echo HACKED\")')"
+    ref_text = ("full_match", "exploit", malicious)
+    result = parse_coordinates(ref_text, 1000, 1000)
+
+    assert result is None, "exec() payload should be rejected"
+    print("PASS: exec() payload is blocked")
+
+
+def test_rce_blocked_eval():
+    """Nested eval() payloads must be blocked."""
+    malicious = "eval('__import__(\"os\").popen(\"id\").read()')"
+    ref_text = ("full_match", "exploit", malicious)
+    result = parse_coordinates(ref_text, 1000, 1000)
+
+    assert result is None, "Nested eval() payload should be rejected"
+    print("PASS: Nested eval() payload is blocked")
+
+
+def test_rce_blocked_lambda():
+    """Lambda-based payloads must be blocked."""
+    malicious = "(lambda: __import__('os').system('echo HACKED'))()"
+    ref_text = ("full_match", "exploit", malicious)
+    result = parse_coordinates(ref_text, 1000, 1000)
+
+    assert result is None, "Lambda payload should be rejected"
+    print("PASS: Lambda payload is blocked")
+
+
+def test_rce_blocked_comprehension():
+    """List comprehension code execution must be blocked."""
+    malicious = "[__import__('os').system('echo HACKED') for x in [1]]"
+    ref_text = ("full_match", "exploit", malicious)
+    result = parse_coordinates(ref_text, 1000, 1000)
+
+    assert result is None, "List comprehension payload should be rejected"
+    print("PASS: List comprehension payload is blocked")
+
+
+if __name__ == "__main__":
+    print("=" * 60)
+    print("Security Regression Tests (OX Security RCE disclosure)")
+    print("=" * 60)
+    print()
+
+    tests = [
+        test_legitimate_coordinates,
+        test_multiple_boxes,
+        test_rce_blocked_import_os,
+        test_rce_blocked_exec,
+        test_rce_blocked_eval,
+        test_rce_blocked_lambda,
+        test_rce_blocked_comprehension,
+    ]
+
+    passed = 0
+    failed = 0
+    for test in tests:
+        try:
+            test()
+            passed += 1
+        except AssertionError as e:
+            print(f"FAIL: {test.__name__}: {e}")
+            failed += 1
+        except Exception as e:
+            print(f"ERROR: {test.__name__}: {e}")
+            failed += 1
+
+    print()
+    print(f"Results: {passed} passed, {failed} failed out of {len(tests)} tests")
+    if failed == 0:
+        print("All security tests passed - RCE vulnerability is patched.")
+    else:
+        print("WARNING: Some tests failed!")

frontend/package.json

@@ -10,6 +10,7 @@
   },
   "dependencies": {
     "axios": "^1.6.5",
+    "dompurify": "^3.3.3",
     "framer-motion": "^11.0.0",
     "lucide-react": "^0.344.0",
     "react": "^18.3.1",

frontend/src/components/ResultPanel.jsx

@@ -2,6 +2,7 @@ import { useEffect, useRef, useState, useCallback } from 'react'
 import { motion, AnimatePresence } from 'framer-motion'
 import { Copy, Download, Sparkles, Loader2, CheckCircle2, ChevronDown } from 'lucide-react'
 import ReactMarkdown from 'react-markdown'
+import DOMPurify from 'dompurify'
 
 export default function ResultPanel({ result, loading, imagePreview, onCopy, onDownload }) {
   const canvasRef = useRef(null)
@@ -230,7 +231,7 @@ export default function ResultPanel({ result, loading, imagePreview, onCopy, onD
               {isHTML ? (
                 <div 
                   className="prose prose-invert prose-sm max-w-none"
-                  dangerouslySetInnerHTML={{ __html: result.text }}
+                  dangerouslySetInnerHTML={{ __html: DOMPurify.sanitize(result.text) }}
                   style={{
                     color: '#e5e7eb',
                   }}