Skip to content

fix: Resolve custom_metadata propagation and LiteLLM thought parts issues #3973

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

Open
sreeharsha1902 wants to merge 9 commits into
base: main
Choose a base branch
from
+11 −3
Open

fix: Resolve custom_metadata propagation and LiteLLM thought parts issues #3973

Show file tree
Hide file tree
Changes from 2 commits
Commits
Show all changes
9 commits
Select commit Hold shift + click to select a range
a7059e3
fix(core): Propagate RunConfig.custom_metadata to Event.custom_metadata
sreeharsha1902 Dec 18, 2025
6b4a3a3
fix(models): Filter thought parts in LiteLLM _get_content()
sreeharsha1902 Dec 18, 2025
19e5aa2
refactor: Use getattr for safer attribute access
sreeharsha1902 Dec 18, 2025
26b665c
chore: Address code review feedback and apply autoformatting
sreeharsha1902 Dec 19, 2025
0fc53fa
Merge branch 'main' into fix/custom-metadata-and-thought-parts
sreeharsha1902 Dec 19, 2025
ee6d643
Merge branch 'main' into fix/custom-metadata-and-thought-parts
sreeharsha1902 Dec 22, 2025
6864401
Merge branch 'main' into fix/custom-metadata-and-thought-parts
sreeharsha1902 Dec 29, 2025
cd0cef7
Merge branch 'main' into fix/custom-metadata-and-thought-parts
sreeharsha1902 Jan 20, 2026
4e80ec7
Merge branch 'main' into fix/custom-metadata-and-thought-parts
sreeharsha1902 Feb 3, 2026
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
2 changes: 2 additions & 0 deletions src/google/adk/flows/llm_flows/base_llm_flow.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -324,6 +324,7 @@ def get_author_for_event(llm_response):
id=Event.new_id(),
invocation_id=invocation_context.invocation_id,
author=get_author_for_event(llm_response),
custom_metadata=invocation_context.run_config.custom_metadata,
Copy link
Copy Markdown
Contributor

@gemini-code-assist gemini-code-assist Bot Dec 18, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

high

The invocation_context.run_config is of type Optional[RunConfig] and could be None. Accessing custom_metadata directly could lead to an AttributeError. It's safer to use getattr to avoid this potential runtime error.

Suggested change
custom_metadata=invocation_context.run_config.custom_metadata,
custom_metadata=getattr(invocation_context.run_config, 'custom_metadata', None),

)

async with Aclosing(
Expand Down Expand Up @@ -438,6 +439,7 @@ async def _run_one_step_async(
invocation_id=invocation_context.invocation_id,
author=invocation_context.agent.name,
branch=invocation_context.branch,
custom_metadata=invocation_context.run_config.custom_metadata,
Copy link
Copy Markdown
Contributor

@gemini-code-assist gemini-code-assist Bot Dec 18, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

high

Similar to the previous comment, invocation_context.run_config can be None. Using getattr provides a safer way to access custom_metadata and prevents a potential AttributeError.

Suggested change
custom_metadata=invocation_context.run_config.custom_metadata,
custom_metadata=getattr(invocation_context.run_config, 'custom_metadata', None),

)
async with Aclosing(
self._call_llm_async(
Expand Down
2 changes: 2 additions & 0 deletions src/google/adk/models/lite_llm.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -533,6 +533,8 @@ async def _get_content(

content_objects = []
for part in parts:
if part.thought:
Copy link
Copy Markdown
Contributor

@gemini-code-assist gemini-code-assist Bot Dec 18, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

medium

The thought attribute on a types.Part object is optional. To prevent a potential AttributeError if a part doesn't have this attribute, it's safer to use getattr(part, 'thought', False) for the check.

Suggested change
if part.thought:
if getattr(part, 'thought', False):

continue
if part.text:
if len(parts) == 1:
return part.text
Expand Down