chore(ci): use GitHub variables for model configuration #1850
+16
−9
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
Replace hardcoded model name and family with GitHub repository variables to improve configuration management and flexibility across workflows.
✅ Deploy Preview for midscene ready!
To edit notification comments on pull requests, go to your Netlify project configuration. |
Deploying midscene with
|
| Latest commit: |
07e088c
|
| Status: | ✅ Deploy successful! |
| Preview URL: | https://5a2d8dfd.midscene.pages.dev |
| Branch Preview URL: | https://chore-use-github-vars-for-mo.midscene.pages.dev |
Replace deprecated `more_actions_needed_by_instruction` field with the new `shouldContinuePlanning` and `finalizeSuccess` fields that were introduced in the planning API refactor. The test now correctly checks: - shouldContinuePlanning is true (task not yet complete) - finalizeSuccess is not true (no successful completion) - log is defined (AI provided preamble message) This aligns with the new complete-task tag mechanism.
The cache-functionality tests were failing with timeouts because: 1. Tests used read-only cache strategy 2. CI environment has no pre-existing cache files 3. read-only mode with empty cache causes matchCache to always return undefined, leading to AI waiting indefinitely until timeout Solution: Use read-write strategy which: - Creates cache if it doesn't exist - Uses existing cache if available - Works correctly in both CI and local environments This fixes test failures in: - should work with explicit cache ID configuration - should work with different cache strategies - should handle cache with performance considerations
The DEBUG=midscene:* environment variable was causing excessive logging (3000+ lines of AI reasoning output), which led to: - Log buffer overflow - Process termination with ELIFECYCLE error - Truncated test output Commenting out DEBUG to allow tests to complete normally. Tests are actually passing - the failure was due to log overflow, not test failures.
yuyutaotao
approved these changes
Jan 23, 2026
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.
Summary
This PR refactors the GitHub Actions workflows to use repository variables instead of hardcoded values for model configuration.
Changes
MIDSCENE_MODEL_NAMEvalue with${{ vars.MIDSCENE_MODEL_NAME }}MIDSCENE_MODEL_FAMILYvalue with${{ vars.MIDSCENE_MODEL_FAMILY }}.github/workflows/ai-evaluation.yml.github/workflows/ai-unit-test.yml.github/workflows/ai.ymlBenefits
MIDSCENE_MODEL_BASE_URL