fix(chat): give extensionless image/audio uploads a valid MIME subtype (#5205)

build_user_content derived the data-URL subtype from the file extension
only (image_format = ext[1:]). An extensionless upload (e.g. a pasted
screenshot) has ext == "", producing "data:image/;base64,..." with an
empty subtype (invalid per RFC 2046) that vision/audio endpoints reject,
silently dropping the attachment. Fall back to the resolved MIME subtype
when the extension is missing; present extensions are unchanged.

routes/chat_routes.py

@@ -876,6 +876,7 @@ def setup_chat_routes(
         # by default without having to send allow_bash in every request.
         if allow_bash is not None and str(allow_bash).lower() != "true":
             disabled_tools.add("bash")
+        _explicit_web_intent = bool(_tool_intent and _tool_intent.category == "web")
         if (
             allow_web_search is not None
             and str(allow_web_search).lower() != "true"

src/document_processor.py

@@ -440,7 +440,10 @@ def build_user_content(
             try:
                 with open(path, "rb") as image_file:
                     encoded_string = base64.b64encode(image_file.read()).decode("utf-8")
-                image_format = ext[1:]
+                # Extensionless uploads (e.g. a pasted screenshot) have no ext,
+                # so fall back to the resolved MIME subtype rather than emitting
+                # an invalid "data:image/;base64," with an empty subtype.
+                image_format = ext[1:] or (mime.split("/", 1)[1] if mime.startswith("image/") else "png")
                 content.append({
                     "type": "image_url",
                     "image_url": {"url": f"data:image/{image_format};base64,{encoded_string}"},
@@ -456,7 +459,7 @@ def build_user_content(
             try:
                 with open(path, "rb") as audio_file:
                     encoded_string = base64.b64encode(audio_file.read()).decode("utf-8")
-                audio_format = ext[1:]
+                audio_format = ext[1:] or (mime.split("/", 1)[1] if mime.startswith("audio/") else "mpeg")
                 content.append({
                     "type": "audio",
                     "audio": {"url": f"data:audio/{audio_format};base64,{encoded_string}"},

tests/test_document_processor_empty_media_subtype.py

@@ -0,0 +1,74 @@
+"""Regression: extensionless image/audio uploads must get a valid MIME subtype.
+
+The data-URL subtype was derived only from the stored file's extension
+(`image_format = ext[1:]`). A pasted screenshot or any file whose stored id
+carries no extension yields `ext == ""`, so the emitted URL was
+`data:image/;base64,...` — an empty MIME subtype (invalid per RFC 2046) that
+vision/audio endpoints reject, silently dropping the attachment. When the
+extension is missing, fall back to the resolved MIME subtype. Extensions that
+are present are unchanged.
+"""
+
+
+class _Handler:
+    def __init__(self, uploads, image=False, audio=False):
+        self.uploads = uploads
+        self._image = image
+        self._audio = audio
+
+    def resolve_upload(self, fid, owner=None):
+        return self.uploads.get(fid)
+
+    def _inside_upload_dir(self, path):
+        return True
+
+    def is_image_file(self, name, mime):
+        return self._image and (mime or "").startswith("image/")
+
+    def is_audio_file(self, name, mime):
+        return self._audio and (mime or "").startswith("audio/")
+
+    def is_document_file(self, name, mime):
+        return False
+
+
+def _blocks(content, block_type):
+    return [b for b in content if isinstance(b, dict) and b.get("type") == block_type]
+
+
+def test_extensionless_image_uses_mime_subtype(tmp_path):
+    import src.document_processor as dp
+
+    p = tmp_path / ("a" * 32)          # bare id, no extension
+    p.write_bytes(b"\x89PNG\r\n\x1a\nfake")
+    uploads = {"img": {"path": str(p), "name": "screenshot", "mime": "image/png"}}
+
+    content = dp.build_user_content("look", ["img"], str(tmp_path), _Handler(uploads, image=True), owner="t")
+    imgs = _blocks(content, "image_url")
+    assert imgs, content
+    assert imgs[0]["image_url"]["url"].startswith("data:image/png;base64,")
+
+
+def test_extensionless_audio_uses_mime_subtype(tmp_path):
+    import src.document_processor as dp
+
+    p = tmp_path / ("b" * 32)
+    p.write_bytes(b"fakeaudio")
+    uploads = {"aud": {"path": str(p), "name": "recording", "mime": "audio/mpeg"}}
+
+    content = dp.build_user_content("listen", ["aud"], str(tmp_path), _Handler(uploads, audio=True), owner="t")
+    auds = _blocks(content, "audio")
+    assert auds, content
+    assert auds[0]["audio"]["url"].startswith("data:audio/mpeg;base64,")
+
+
+def test_extension_present_is_unchanged(tmp_path):
+    import src.document_processor as dp
+
+    p = tmp_path / "pic.png"
+    p.write_bytes(b"\x89PNG\r\n\x1a\n")
+    uploads = {"img": {"path": str(p), "name": "pic.png", "mime": "image/png"}}
+
+    content = dp.build_user_content("look", ["img"], str(tmp_path), _Handler(uploads, image=True), owner="t")
+    imgs = _blocks(content, "image_url")
+    assert imgs[0]["image_url"]["url"].startswith("data:image/png;base64,")