modelcontextprotocol/python-sdk
Watch
1
0
Fork
You've already forked python-sdk
 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

MCPServer: content-block returns are unstructured, prompt messages take Image/Audio (#3320)

Browse source
This commit is contained in:
Max 2026-08-17 14:12:31 +01:00 • committed by GitHub
commit fb443cc4b2
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
1 changed files with 290 additions and 56 deletions
Download patch file Download diff file Expand all files Collapse all files

4
src/mcp/server/mcpserver/server.py
View file

@ -918,8 +918,8 @@ class MCPServer(Generic[LifespanResultT]):
) -> Callable[[_CallableT], _CallableT]:
"""Decorator to register a prompt.
The function returns the prompt messages (a string, `Message`, dict,
or a sequence of these), or an `InputRequiredResult` to request
The function returns the prompt messages (a string, content block, `Image`/`Audio`,
`Message`, dict, or a sequence of these), or an `InputRequiredResult` to request
client input first (the 2026-07-28 multi-round-trip flow — read
`ctx.input_responses` on the retry).