[OPIK-5242] [OPIK-5248] [FE] fix: evaluation suite save failures and Result column truncation by alexkuzmik · Pull Request #5844 · comet-ml/opik

alexkuzmik · 2026-03-25T11:47:45Z

Details

Fixes three bugs in the evaluation suite UI:

OPIK-5242 Case 1 — Editing a suite created with only a name: buildPayload threw because suite.latest_version was undefined. Now detects the no-version state and performs a two-step save: creates an initial metadata-only version (evaluators + execution policy) first, then applies item changes using the returned version ID as base.

OPIK-5242 Case 2 — Creating a suite with items + assertions/policy at once: Items were uploaded before evaluation criteria, so by the time applyEvaluationCriteria ran with base_version: null, the dataset already had a version (created by the item upload), causing a backend 400 rejection. Swapped the order to match the Python SDK: apply criteria first (creates the initial version), then upload items.

OPIK-5248 — Result column truncated when all columns enabled: The pinned "Result" column in the experiment items table had no minSize, allowing it to be compressed to unreadable widths. Added minSize: 140 to match the column's default size.

Change checklist

User facing
Documentation update

Issues

OPIK-5242
OPIK-5248

AI-WATERMARK

AI-WATERMARK: yes

If yes:
- Tools: Claude Code (VS Code extension)
- Model(s): Claude Opus 4.6
- Scope: Bug analysis, fix implementation, PR creation
- Human verification: Code review of all changes before commit

Testing

TypeScript type check passes: npx tsc --noEmit — no errors
ESLint passes on all 4 changed files
Manual testing scenarios for OPIK-5242:
1. Create evaluation suite with just a name → edit to add items/assertions/policy → Save changes → should succeed (two-step save)
2. Create evaluation suite with name + CSV items + assertions + execution policy → should save everything (criteria applied before items)
3. Create evaluation suite with name + assertions/policy but no CSV → should save criteria (no change from working path)
4. Edit existing suite with version → Save changes → should work as before (single-step save, no regression)
Manual testing for OPIK-5248:
1. Open experiment items tab for an evaluation suite experiment
2. Enable all columns via the columns button
3. Verify the "Result" column remains readable (shows "Passed"/"Failed" fully)

Documentation

N/A — Bug fixes only, no new configuration or features.

…ersion and creation flows Fix two bugs in evaluation suite saving: 1. Editing a suite created with only a name would fail because buildPayload threw when latest_version was missing. Now handles no-version suites with a two-step save: create metadata version first, then apply item changes. 2. Creating a suite with items + assertions/policy failed because items were uploaded before criteria, causing the backend to reject base_version=null on a dataset that already had versions. Swapped the order to match the Python SDK: apply criteria first, then upload items. Implements OPIK-5242: [Evaluation suite] - issues during creating evaluation suite via UI Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

baz-reviewer · 2026-03-25T11:52:29Z

apps/opik-frontend/src/v2/pages/EvaluationSuiteItemsPage/EvaluationSuiteItemsPage.tsx

  const handleSaveChanges = (tags?: string[], changeDescription?: string) => {
    if (changesMutation.isPending) return;

+    if (hasNoVersion) {
+      // Two-step save: create initial metadata version first, then apply
+      // item changes on top of it. The backend requires base_version=null
+      // with no items for the first version.
+      changesMutation.mutate(
+        buildInitialVersionPayload({ tags, changeDescription }),
+        {
+          onSuccess: (initialVersion) => {
+            const itemPayload = buildPayload({


When hasNoVersion the second buildPayload call omits tags/changeDescription — should we pass them so the final version preserves user metadata?

_{Finding type: Logical Bugs | Severity: 🔴 High}

Want Baz to fix this for you? Activate Fixer

Other fix methods

Prompt for AI Agents:

Before applying, verify this suggestion against the current code. In apps/opik-frontend/src/v2/pages/EvaluationSuiteItemsPage/EvaluationSuiteItemsPage.tsx around lines 200 to 225, the handleSaveChanges function (the hasNoVersion branch) builds an itemPayload with buildPayload({ baseVersionOverride: initialVersion?.id }) but does not pass the tags and changeDescription from the Add Version dialog. Refactor this by calling buildPayload({ baseVersionOverride: initialVersion?.id, tags, changeDescription }) so the second mutation includes the same metadata, ensuring the final version preserves tags and changeDescription provided by the user. Keep the existing conflict/onError handling the same.

Commit 4864e55 addressed this comment by passing the tags and changeDescription into the second buildPayload call in the hasNoVersion flow, so the final version preserves the metadata from the Add Version dialog and the existing conflict handling remains unchanged.

github-actions · 2026-03-25T12:12:06Z

🔄 Test environment deployment process has started

Phase 1: Deploying base version 1.10.49 (from main branch) if environment doesn't exist
Phase 2: Building new images from PR branch aliaksandrk/OPIK-5242-evaluation-suite-save-bugs
Phase 3: Will deploy newly built version after build completes

You can monitor the progress here.

…dPayload call Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

alexkuzmik · 2026-03-25T12:13:21Z

Re: baz-reviewer comment about missing tags/changeDescription in the second buildPayload call —

Fixed in 4864e55 — now passing tags and changeDescription to the second buildPayload call so the final version preserves user metadata.

🤖 Reply posted via /address-github-pr-comments

baz-reviewer · 2026-03-25T12:13:45Z

Got it, and thanks for addressing that—I'll save this to memory after the PR is merged so I can better recognize this pattern in future reviews.

github-actions · 2026-03-25T12:24:12Z

🔄 Test environment deployment process has started

Phase 1: Deploying base version 1.10.49 (from main branch) if environment doesn't exist
Phase 2: Building new images from PR branch aliaksandrk/OPIK-5242-evaluation-suite-save-bugs
Phase 3: Will deploy newly built version after build completes

You can monitor the progress here.

CometActions · 2026-03-25T12:31:48Z

✅ Test environment is now available!

To configure additional Environment variables for your environment, run [Deploy Opik AdHoc Environment workflow] (https://github.com/comet-ml/comet-deployment/actions/workflows/deploy_opik_adhoc_env.yaml)

Access Information

URL: https://pr-5844.dev.comet.com
Cluster: comet-ml-development
Namespace: pr-5844
Version: 1.10.49-5844-merge-1672
Application logs: View in Grafana

The deployment has completed successfully and the version has been verified.

…columns are enabled Add minSize: 140 to the pinned Result column so it cannot be compressed below its content width when all columns are visible. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

github-actions · 2026-03-25T13:23:59Z

🔄 Test environment deployment process has started

Phase 1: Deploying base version 1.10.49 (from main branch) if environment doesn't exist
Phase 2: Building new images from PR branch aliaksandrk/OPIK-5242-evaluation-suite-save-bugs
Phase 3: Will deploy newly built version after build completes

You can monitor the progress here.

CometActions · 2026-03-25T13:29:41Z

❌ Test environment deployment failed

The deployment encountered an error. Please check the deployment logs for details.

CometActions · 2026-03-25T13:29:42Z

❌ Test environment deployment failed

The deployment encountered an error. Please check the deployment logs for details.

CometActions · 2026-03-25T13:31:50Z

✅ Test environment is now available!

To configure additional Environment variables for your environment, run [Deploy Opik AdHoc Environment workflow] (https://github.com/comet-ml/comet-deployment/actions/workflows/deploy_opik_adhoc_env.yaml)

Access Information

URL: https://pr-5844.dev.comet.com
Cluster: comet-ml-development
Namespace: pr-5844
Version: 1.10.49-5844-merge-1677
Application logs: View in Grafana

The deployment has completed successfully and the version has been verified.

CometActions · 2026-03-26T04:53:52Z

🌙 Nightly cleanup: The test environment for this PR (pr-5844) has been cleaned up to free cluster resources. PVCs are preserved — re-deploy to restore the environment.

awkoy

LGTM!

CometActions · 2026-03-27T04:53:00Z

🌙 Nightly cleanup: The test environment for this PR (pr-5844) has been cleaned up to free cluster resources. PVCs are preserved — re-deploy to restore the environment.

CometActions · 2026-03-28T04:42:34Z

🌙 Nightly cleanup: The test environment for this PR (pr-5844) has been cleaned up to free cluster resources. PVCs are preserved — re-deploy to restore the environment.

CometActions · 2026-03-29T04:56:24Z

🌙 Nightly cleanup: The test environment for this PR (pr-5844) has been cleaned up to free cluster resources. PVCs are preserved — re-deploy to restore the environment.

github-actions bot added Frontend typescript *.ts *.tsx labels Mar 25, 2026

github-actions bot assigned alexkuzmik Mar 25, 2026

baz-reviewer bot reviewed Mar 25, 2026

View reviewed changes

alexkuzmik added the test-environment Deploy Opik adhoc environment label Mar 25, 2026

alexkuzmik removed the test-environment Deploy Opik adhoc environment label Mar 25, 2026

fix(evaluation-suite): pass tags and changeDescription to second buil…

4864e55

…dPayload call Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

alexkuzmik added the test-environment Deploy Opik adhoc environment label Mar 25, 2026

[OPIK-5248] [FE] fix: prevent Result column from truncating when all …

53dc190

…columns are enabled Add minSize: 140 to the pinned Result column so it cannot be compressed below its content width when all columns are visible. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

alexkuzmik changed the title ~~[OPIK-5242] [FE] fix: resolve evaluation suite save failures for no-version and creation flows~~ [OPIK-5242] [OPIK-5248] [FE] fix: evaluation suite save failures and Result column truncation Mar 25, 2026

alexkuzmik added test-environment Deploy Opik adhoc environment and removed test-environment Deploy Opik adhoc environment labels Mar 25, 2026

comet-ml deleted a comment from github-actions bot Mar 25, 2026

alexkuzmik marked this pull request as ready for review March 25, 2026 13:55

alexkuzmik requested a review from a team as a code owner March 25, 2026 13:55

CometActions removed the test-environment Deploy Opik adhoc environment label Mar 26, 2026

awkoy approved these changes Mar 26, 2026

View reviewed changes

Conversation

alexkuzmik commented Mar 25, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Details

Change checklist

Issues

AI-WATERMARK

Testing

Documentation

Uh oh!

baz-reviewer bot Mar 25, 2026

Choose a reason for hiding this comment

Uh oh!

baz-reviewer bot Mar 25, 2026

Choose a reason for hiding this comment

Uh oh!

github-actions bot commented Mar 25, 2026

Uh oh!

alexkuzmik commented Mar 25, 2026

Uh oh!

baz-reviewer bot commented Mar 25, 2026

Uh oh!

github-actions bot commented Mar 25, 2026

Uh oh!

CometActions commented Mar 25, 2026

Access Information

Uh oh!

github-actions bot commented Mar 25, 2026

Uh oh!

CometActions commented Mar 25, 2026

Uh oh!

CometActions commented Mar 25, 2026

Uh oh!

CometActions commented Mar 25, 2026

Access Information

Uh oh!

CometActions commented Mar 26, 2026

Uh oh!

awkoy left a comment

Choose a reason for hiding this comment

Uh oh!

CometActions commented Mar 27, 2026

Uh oh!

CometActions commented Mar 28, 2026

Uh oh!

CometActions commented Mar 29, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

alexkuzmik commented Mar 25, 2026 •

edited

Loading