🤖 fix: make OpenAI truncation integration test more robust #253
Conversation
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
github-merge-queue
bot
removed this pull request from the merge queue due to failed status checks
The integration test was flaky because AI models sometimes complete tool calls without generating text output, causing the stream to never emit stream-end. Fix: Modified test prompt to request confirmation after tool execution. This encourages the AI to generate text output, ensuring the stream completes properly. Added analysis document explaining the root cause and potential solutions.
ammar-agent
force-pushed
the
investigate-e2e-flake
branch
from
42c093d to
d87478d
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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Outdated suggestions cannot be applied.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.
Problem
Integration test was flaking:
OpenAI auto truncation integration > should include full file_edit diff in UI/history but redact it from the next provider requestFailure mode: Stream completes tool calls but never emits
stream-endevent, causing test timeout.Root cause: AI models can complete tool execution without generating text output. This is non-deterministic behavior - sometimes the model responds with text after tools, sometimes it doesn't.
See full analysis in
E2E_FLAKE_ANALYSIS.md.Solution
Modified test prompt to explicitly request confirmation after tool execution:
This encourages the AI to generate text output after completing tools, ensuring the stream finishes properly.
Trade-offs
Short-term fix: Prompt modification reduces flakiness significantly
Long-term fix: Stream manager should detect tool-only responses and auto-emit stream-end (tracked for future work)
Testing
Generated with
cmux