Skip to content

docs: fix fulfillment typos in session runner #1795

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
dtmeadows-ant merged 1 commit into from
Aug 7, 2026
+4 −4
Merged

docs: fix fulfillment typos in session runner #1795

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 3 additions & 3 deletions src/anthropic/lib/tools/_beta_session_runner.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -270,7 +270,7 @@ class DispatchedToolCall:
want to surface that or retry at a higher level — and also ``False``, with
``result`` left ``None``, when the tool name is not one this runner owns and
it deliberately posted nothing, leaving the ``tool_use_id`` pending for its
owner (the split-client partial-fulfilment behavior), or when the call was
owner (the split-client partial-fulfillment behavior), or when the call was
denied and never executed (see ``confirmation``)."""

confirmation: Literal["allow", "deny"] | None = None
Expand Down Expand Up @@ -382,7 +382,7 @@ class SessionToolRunner:
A self-hosted session is commonly serviced by **two** clients at once: this
runner inside the customer's sandbox (registered with the file/shell sandbox
tools) and the customer's app backend (handling the agent's ``custom``
function tools). The Sessions API has a partial-fulfilment contract: when a
function tools). The Sessions API has a partial-fulfillment contract: when a
session pauses on ``requires_action`` the pending tool-call ids can mix both
kinds, and each client must post results **only** for the ids it owns and
leave the rest pending for the other client. A tool-call event whose name is
Expand Down Expand Up @@ -833,7 +833,7 @@ async def _execute(self, ev: DispatchedToolUseEvent, confirmation: Literal["allo
tool = self._tools_by_name.get(ev.name)
tool_result: DispatchedToolResultParams | None
if tool is None:
# Skip unowned (split-client partial fulfilment): a name this
# Skip unowned (split-client partial fulfillment): a name this
# runner is not registered for belongs to the other client
# servicing this session (typically the customer's app backend
# handling custom tools). Post NO result, do not mark it answered,
Expand Down
2 changes: 1 addition & 1 deletion tests/lib/tools/test_session_runner.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -472,7 +472,7 @@ async def weather(_input: dict[str, Any]) -> str:
assert items == []


# ---------- skip unowned tools (split-client partial fulfilment) -----------
# ---------- skip unowned tools (split-client partial fulfillment) -----------


@pytest.mark.asyncio()
Expand Down
Loading