fix(eval): ignore UI-only InvocationEvent fields on eval case save - #7096
Closed
Anusha0501 wants to merge 2 commits into
Closed
Anusha0501 wants to merge 2 commits into
Anusha0501 wants to merge 2 commits into
Conversation
|
Thanks for your pull request! It looks like this may be your first contribution to a Google open source project. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA). View this failed invocation of the CLA check for more information. For the most up to date status, view the checks section at the bottom of the pull request. |
Anusha0501
force-pushed
the
fix/eval-ignore-web-ui-invocation-indices
branch
from
September 12, 2026 08:54
7d82908 to
cecec57
Compare
The adk web eval editor attaches invocationIndex and toolUseIndex to each event. Those fields are not part of the eval schema, so PUT /eval-cases returned 422. Ignore unknown fields so the save succeeds without persisting the UI indices. Fixes google#7019
Anusha0501
force-pushed
the
fix/eval-ignore-web-ui-invocation-indices
branch
from
September 12, 2026 08:57
cecec57 to
a553c55
Compare
copybara-service Bot
pushed a commit
that referenced
this pull request
Sep 16, 2026
Merge #7096 **Please ensure you have read the [contribution guide](https://github.com/google/adk-python/blob/main/CONTRIBUTING.md) before creating a pull request.** ### Link to Issue or Description of Change **1. Link to an existing issue (if applicable):** - Closes: #7019 - Related: #7026 (different 422: missing `RubricContent.text_property` default; does not cover this extra-fields case) **2. Or, if no issue exists, describe the change:** **Problem:** Saving an eval case from the `adk web` editor returns HTTP 422 Unprocessable Entity. The editor serializes UI-only transcript fields (`invocationIndex`, `toolUseIndex`) onto each `InvocationEvent`. `EvalBaseModel` uses `extra="forbid"`, so FastAPI rejects the PUT body as `extra_forbidden`. **Solution:** Override `InvocationEvent` with `extra="ignore"` (same ConfigDict pattern as `EvalCase` / `SessionInput`). Unknown UI fields are dropped at parse time, the save succeeds, and those indices are not persisted in the eval-case JSON. ### Testing Plan **Unit Tests:** - [x] I have added or updated unit tests for my change. - [x] All unit tests pass locally. `PYTHONPATH=src pytest tests/unittests/evaluation/test_eval_case.py` ``` 23 passed ``` Added `test_eval_case_put_accepts_web_ui_transcript_indices`, which validates the reporter payload from #7019 and asserts `invocationIndex` / `toolUseIndex` are not stored. **Manual End-to-End (E2E) Tests:** - Run `adk web` for an agent that uses tools - Create a session, add it to an eval set, edit a message, click Save - Expect HTTP 200 (not 422) - Confirm the stored eval-case JSON does not contain `invocationIndex` or `toolUseIndex` ### Checklist - [x] I have read the [CONTRIBUTING.md](https://github.com/google/adk-python/blob/main/CONTRIBUTING.md) document. - [x] I have performed a self-review of my own code. - [x] I have commented my code, particularly in hard-to-understand areas. - [x] I have added tests that prove my fix is effective or that my feature works. - [x] New and existing unit tests pass locally with my changes. - [ ] I have manually tested my changes end-to-end. - [ ] Any dependent changes have been merged and published in downstream modules. ### Additional context `#7026` only defaults `RubricContent.text_property`. The reporter confirmed that does not fix this save. Maintainer diagnosis on #7019: `extra_forbidden` on `invocationIndex` and `toolUseIndex` inside `conversation[0].intermediateData.invocationEvents`. Co-authored-by: Yi Liu <yiliuly@google.com> COPYBARA_INTEGRATE_REVIEW=#7096 from Anusha0501:fix/eval-ignore-web-ui-invocation-indices 782eb0c PiperOrigin-RevId: 982699482
Collaborator
|
Thank you @Anusha0501 for your contribution! 🎉 Your changes have been successfully imported and merged via Copybara in commit a4a50a4. Closing this PR as the changes are now in the main branch. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Please ensure you have read the contribution guide before creating a pull request.
Link to Issue or Description of Change
1. Link to an existing issue (if applicable):
RubricContent.text_propertydefault; does not cover this extra-fields case)2. Or, if no issue exists, describe the change:
Problem:
Saving an eval case from the
adk webeditor returns HTTP 422 Unprocessable Entity. The editor serializes UI-only transcript fields (invocationIndex,toolUseIndex) onto eachInvocationEvent.EvalBaseModelusesextra="forbid", so FastAPI rejects the PUT body asextra_forbidden.Solution:
Override
InvocationEventwithextra="ignore"(same ConfigDict pattern asEvalCase/SessionInput). Unknown UI fields are dropped at parse time, the save succeeds, and those indices are not persisted in the eval-case JSON.Testing Plan
Unit Tests:
PYTHONPATH=src pytest tests/unittests/evaluation/test_eval_case.pyAdded
test_eval_case_put_accepts_web_ui_transcript_indices, which validates the reporter payload from #7019 and assertsinvocationIndex/toolUseIndexare not stored.Manual End-to-End (E2E) Tests:
adk webfor an agent that uses toolsinvocationIndexortoolUseIndexChecklist
Additional context
#7026only defaultsRubricContent.text_property. The reporter confirmed that does not fix this save. Maintainer diagnosis on #7019:extra_forbiddenoninvocationIndexandtoolUseIndexinsideconversation[0].intermediateData.invocationEvents.