[nborges-aws]

Description

The config bundle TUI currently is unable to render a bundle which was not created with the default 'mainline' branch. The TUI always expects mainline, thus a bundle with a custom branch would fail to render, and show the user an error.

This PR adds a branch selection screen to the config bundle selection flow. After selecting a bundle, the user is prompted to select the branch they intend to view. This follows the pattern established for runtime-endpoint qualifier, and config-bundle version.

The follow video shows:

• TUI on the current vended RC of v1 CLI: agentcore ... command, fails to load bundle
• TUI on my local dev build of v1 CLI: ac-dev ... user is prompted to select branch, and bundle is rendered

Screen.Recording.2026-10-20.at.3.17.04.PM.mov

Type of Change

• Bug fix
• New feature
• Breaking change
• Documentation update
• Other (please describe):

Testing

How have you tested the change?

• bun run test (4698 pass, 0 fail)
• I ran npm run test:unit and npm run test:integ
• I ran npm run typecheck
• I ran npm run lint
• If I modified src/assets/, I ran npm run test:update-snapshots and committed the updated snapshots

Checklist

• I have read the CONTRIBUTING document
• I have added any necessary tests that prove my fix is effective or my feature works
• I have updated the documentation accordingly
• I have added an appropriate example to the documentation to outline the feature, or no new docs are needed
• My changes generate no new warnings
• Any dependent changes have been merged and published

---

By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the
terms of your choice.

[agentcore-devx-automation]

Claude Security Review: no high-confidence findings. (run)

[agentcore-devx-automation]

AgentCore Harness Review

Verdict: Looks good

Nice refactor — inserting a branch picker between the bundle list and the detail view, with the new latestPerBranch filter driving it, hangs together cleanly.

A few things I checked that are fine as-is:

• Route specificity in Root.tsx: the static-segment route get/:bundleId/branch/:branchName is declared before get/:bundleId/:versionId, and React Router v6 scores the static segment higher regardless of declaration order, so a versionId colliding with the literal "branch" would still route correctly.
• getConfigurationBundle with a versionId: branchName is ignored by the client in that branch, so defaulting to "mainline" in screen.tsx when versionId is present is harmless.
• Query key disambiguation: adding the "branch" / "version" discriminator to the useQuery key avoids cross-contamination between the two modes. Good.
• Testing style: tests use the real TestCoreClient and only stub at the SDK boundary — consistent with the project's "no excessive mocking" guidance.
• Telemetry: none added, but no sibling eval screens instrument telemetry today, so this isn't a regression.

No blocking issues from me.

[codecov-commenter]

Codecov Report

❌ Patch coverage is 98.96907% with 1 line in your changes missing coverage. Please review.
✅ Project coverage is 96.62%. Comparing base (57a3b3b) to head (e1473f9).

Files with missing lines	Patch %	Lines
src/components/ConfigBundleBranchPicker.tsx	98.30%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff            @@
##           refactor    #2627   +/-   ##
=========================================
  Coverage     96.61%   96.62%           
=========================================
  Files           704      706    +2     
  Lines         58404    58494   +90     
=========================================
+ Hits          56429    56519   +90     
  Misses         1975     1975           

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[jariy17]

LGTM

[agentcore-devx-automation]

Claude Security Review: no high-confidence findings. (run)