Prompt functions may return bare content blocks, Image or Audio

render() special-cased str and JSON-dumped anything else that was not a Message or
dict, so a prompt returning Image(...) or a ready-made content block (or a list mixing
captions and images) reached the client as the object's repr or a JSON blob. Bare
content now becomes one user message via UserMessage(msg), making Message.__init__
the single place prompt content is coerced; the JSON-dump fallback for other values is
unchanged. SyncPromptResult is widened to match.

Author: maxisbey

src/mcp/server/mcpserver/prompts/base.py

@@ -62,7 +62,8 @@ def __init__(self, content: str | ContentBlock | Image | Audio, **kwargs: Any):
 
 message_validator = TypeAdapter[UserMessage | AssistantMessage](UserMessage | AssistantMessage)
 
-SyncPromptResult = str | Message | dict[str, Any] | InputRequiredResult | Sequence[str | Message | dict[str, Any]]
+_PromptResultItem = str | ContentBlock | Image | Audio | Message | dict[str, Any]
+SyncPromptResult = _PromptResultItem | InputRequiredResult | Sequence[_PromptResultItem]
 PromptResult = SyncPromptResult | Awaitable[SyncPromptResult]
 
 
@@ -98,7 +99,7 @@ def from_function(
         """Create a Prompt from a function.
 
         The function can return:
-        - A string (converted to a message)
+        - A string, content block, `Image` or `Audio` (each becomes a user message)
         - A Message object
         - A dict (converted to a message)
         - A sequence of any of the above
@@ -192,9 +193,8 @@ async def render(
                         messages.append(msg)
                     elif isinstance(msg, dict):
                         messages.append(message_validator.validate_python(msg))
-                    elif isinstance(msg, str):
-                        content = TextContent(type="text", text=msg)
-                        messages.append(UserMessage(content=content))
+                    elif isinstance(msg, str | ContentBlock | Image | Audio):  # bare content is one user message
+                        messages.append(UserMessage(msg))
                     else:  # pragma: no cover
                         content = pydantic_core.to_json(msg, fallback=str, indent=2).decode()
                         messages.append(Message(role="user", content=content))

tests/server/mcpserver/prompts/test_base.py

@@ -290,3 +290,29 @@ def deck(topic: str) -> list[_Slide]:
 
     [listed] = await mcp.list_prompts()
     assert [arg.name for arg in listed.arguments or []] == ["topic"]
+
+
+_PNG = Image(data=b"img", format="png")
+_PNG_BLOCK = ImageContent(type="image", data="aW1n", mime_type="image/png")
+_DOC = EmbeddedResource(
+    type="resource", resource=TextResourceContents(uri="file://notes.md", text="notes", mime_type="text/markdown")
+)
+
+
+@pytest.mark.anyio
+@pytest.mark.parametrize(
+    ("returned", "expected"),
+    [
+        (_PNG, [UserMessage(_PNG_BLOCK)]),
+        (_DOC, [UserMessage(_DOC)]),
+        (["Look at this:", _PNG], [UserMessage("Look at this:"), UserMessage(_PNG_BLOCK)]),
+    ],
+)
+async def test_bare_content_returned_from_a_prompt_becomes_user_messages(returned: Any, expected: list[Message]):
+    """SDK-defined: what a tool may return bare (a content block, `Image`, `Audio`), a prompt may too;
+    each item becomes one user message instead of being JSON-dumped into text."""
+
+    def fn() -> Any:
+        return returned
+
+    assert await Prompt.from_function(fn).render(None, Context()) == expected