Fix: stored XSS via HTML File upload and inline Rendering in file get (#13202)

### What problem does this PR solve?

Fix stored XSS via HTML file upload and inline rendering in
/v1/file/get/<id>

### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)

api/apps/document_app.py

@@ -44,7 +44,7 @@ from api.utils.api_utils import (
 from api.utils.file_utils import filename_type, thumbnail
 from common.file_utils import get_project_base_directory
 from common.constants import RetCode, VALID_TASK_STATUS, ParserType, TaskStatus
-from api.utils.web_utils import CONTENT_TYPE_MAP, html2pdf, is_valid_url
+from api.utils.web_utils import CONTENT_TYPE_MAP, apply_safe_file_response_headers, html2pdf, is_valid_url
 from deepdoc.parser.html_parser import RAGFlowHtmlParser
 from rag.nlp import search, rag_tokenizer
 from common import settings
@@ -718,13 +718,11 @@ async def get(doc_id):
 
         ext = re.search(r"\.([^.]+)$", doc.name.lower())
         ext = ext.group(1) if ext else None
+        content_type = None
         if ext:
-            if doc.type == FileType.VISUAL.value:
-
-                content_type = CONTENT_TYPE_MAP.get(ext, f"image/{ext}")
-            else:
-                content_type = CONTENT_TYPE_MAP.get(ext, f"application/{ext}")
-            response.headers.set("Content-Type", content_type)
+            fallback_prefix = "image" if doc.type == FileType.VISUAL.value else "application"
+            content_type = CONTENT_TYPE_MAP.get(ext, f"{fallback_prefix}/{ext}")
+        apply_safe_file_response_headers(response, content_type, ext)
         return response
     except Exception as e:
         return server_error_response(e)
@@ -737,7 +735,8 @@ async def download_attachment(attachment_id):
         ext = request.args.get("ext", "markdown")
         data = await thread_pool_exec(settings.STORAGE_IMPL.get, current_user.id, attachment_id)
         response = await make_response(data)
-        response.headers.set("Content-Type", CONTENT_TYPE_MAP.get(ext, f"application/{ext}"))
+        content_type = CONTENT_TYPE_MAP.get(ext, f"application/{ext}")
+        apply_safe_file_response_headers(response, content_type, ext)
 
         return response
 

api/apps/file_app.py

@@ -31,7 +31,7 @@ from api.db.services import duplicate_name
 from api.db.services.file_service import FileService
 from api.utils.api_utils import get_json_result, get_request_json
 from api.utils.file_utils import filename_type
-from api.utils.web_utils import CONTENT_TYPE_MAP
+from api.utils.web_utils import CONTENT_TYPE_MAP, apply_safe_file_response_headers
 from common import settings
 
 @manager.route('/upload', methods=['POST'])  # noqa: F821
@@ -364,12 +364,11 @@ async def get(file_id):
         response = await make_response(blob)
         ext = re.search(r"\.([^.]+)$", file.name.lower())
         ext = ext.group(1) if ext else None
+        content_type = None
         if ext:
-            if file.type == FileType.VISUAL.value:
-                content_type = CONTENT_TYPE_MAP.get(ext, f"image/{ext}")
-            else:
-                content_type = CONTENT_TYPE_MAP.get(ext, f"application/{ext}")
-            response.headers.set("Content-Type", content_type)
+            fallback_prefix = "image" if file.type == FileType.VISUAL.value else "application"
+            content_type = CONTENT_TYPE_MAP.get(ext, f"{fallback_prefix}/{ext}")
+        apply_safe_file_response_headers(response, content_type, ext)
         return response
     except Exception as e:
         return server_error_response(e)

api/apps/sdk/files.py

@@ -28,7 +28,7 @@ from api.db import FileType
 from api.db.services import duplicate_name
 from api.db.services.file_service import FileService
 from api.utils.file_utils import filename_type
-from api.utils.web_utils import CONTENT_TYPE_MAP
+from api.utils.web_utils import CONTENT_TYPE_MAP, apply_safe_file_response_headers
 from common import settings
 from common.constants import RetCode
 
@@ -623,11 +623,12 @@ async def get(tenant_id, file_id):
 
         response = await make_response(blob)
         ext = re.search(r"\.([^.]+)$", file.name)
-        if ext:
-            if file.type == FileType.VISUAL.value:
-                response.headers.set('Content-Type', 'image/%s' % ext.group(1))
-            else:
-                response.headers.set('Content-Type', 'application/%s' % ext.group(1))
+        extension = ext.group(1).lower() if ext else None
+        content_type = None
+        if extension:
+            fallback_prefix = "image" if file.type == FileType.VISUAL.value else "application"
+            content_type = CONTENT_TYPE_MAP.get(extension, f"{fallback_prefix}/{extension}")
+        apply_safe_file_response_headers(response, content_type, extension)
         return response
     except Exception as e:
         return server_error_response(e)
@@ -640,7 +641,8 @@ async def download_attachment(tenant_id, attachment_id):
         ext = request.args.get("ext", "markdown")
         data = await thread_pool_exec(settings.STORAGE_IMPL.get, tenant_id, attachment_id)
         response = await make_response(data)
-        response.headers.set("Content-Type", CONTENT_TYPE_MAP.get(ext, f"application/{ext}"))
+        content_type = CONTENT_TYPE_MAP.get(ext, f"application/{ext}")
+        apply_safe_file_response_headers(response, content_type, ext)
 
         return response
 

api/utils/web_utils.py

@@ -92,6 +92,46 @@ CONTENT_TYPE_MAP = {
 }
 
 
+FORCE_ATTACHMENT_EXTENSIONS = {
+    "htm",
+    "html",
+    "shtml",
+    "xht",
+    "xhtml",
+    "xml",
+    "mhtml",
+    "svg",
+}
+
+
+FORCE_ATTACHMENT_CONTENT_TYPES = {
+    "text/html",
+    "image/svg+xml",
+    "application/xhtml+xml",
+    "text/xml",
+    "application/xml",
+    "multipart/related",
+}
+
+
+def should_force_attachment(ext: str | None, content_type: str | None = None) -> bool:
+    normalized_ext = (ext or "").lower().strip(".")
+    if normalized_ext in FORCE_ATTACHMENT_EXTENSIONS:
+        return True
+    normalized_type = (content_type or "").lower()
+    return normalized_type in FORCE_ATTACHMENT_CONTENT_TYPES
+
+
+def apply_safe_file_response_headers(response, content_type: str | None, ext: str | None = None):
+    if content_type:
+        response.headers.set("Content-Type", content_type)
+    force_attachment = should_force_attachment(ext, content_type)
+    if force_attachment:
+        response.headers.set("X-Content-Type-Options", "nosniff")
+        response.headers.set("Content-Disposition", "attachment")
+    return response
+
+
 def html2pdf(
     source: str,
     timeout: int = 2,
@@ -188,10 +228,9 @@ def get_float(req: dict, key: str, default: float | int = 10.0) -> float:
         return parsed if parsed > 0 else default
     except (TypeError, ValueError):
         return default
-    
+
 
 async def send_email_html(to_email: str, subject: str, template_key: str, **context):
-
     body = await render_template_string(EMAIL_TEMPLATES.get(template_key), **context)
     msg = MIMEText(body, "plain", "utf-8")
     msg["Subject"] = Header(subject, "utf-8")
@@ -236,10 +275,10 @@ def otp_keys(email: str):
 
 def hash_code(code: str, salt: bytes) -> str:
     import hashlib
-    import hmac 
+    import hmac
+
     return hmac.new(salt, (code or "").encode("utf-8"), hashlib.sha256).hexdigest()
-    
+
 
 def captcha_key(email: str) -> str:
     return f"captcha:{email}"
-