[Aaryan430]

Stacked on #496 (which is stacked on #494). GitHub can't stack a PR on a fork branch, so the earlier commits here are theirs. Review only the last commit, 1643f4a. It does not depend on #498.

An agent that fetched an SDK build's accessibility issues right after the test run reported 14 issues as final ("✅ All issues retrieved"). The build completed minutes later with 27. After the tests end, a build stays finalizing, which on an advanced scan is the dashboard's "advanced issue processing", and the export serves whatever exists so far.

Recorded on a fresh build, polling every 5 s:

Time	Build status (list_builds)	Issues the API returns
tests running	running	0
tests end	finalizing	14
~3 min later	completed	27

Change

fetchAccessibilityIssues reads the report's status before reading its issues:

Source	Where the status comes from
Website Scanner run	GET …/scans/{id}/scan_runs/{run}/status
Workflow Analyzer report	GET …/reports/{id} → summary.status
Automated Tests build	its project's list_builds row (the build overview has no status), found by id in the newest builds

• While the report is processing (queued / running / processing / finalizing), a first read waits: it polls every 10 s for up to 5 min, the same budget startAccessibilityScan uses, and sends progress notifications. Then it reads the issues and says it waited.
• If processing outlasts the wait, the answer starts with a warning that these are only the issues found so far, the total will grow, and to fetch again. It ends with "All issues found so far retrieved; the report is still processing" instead of "✅ All issues retrieved".
• A later page doesn't wait again but still warns. A status that can't be read is not waited on (no behaviour change).

Verification

Ran a fresh SDK test (accessibility + advanced on) and called this branch's build over MCP the moment the run ended:

  +1s progress: The report is still finalizing; waiting for it to complete (1s of up to 5m 00s).
  ...
+133s progress: The report is still finalizing; waiting for it to complete (2m 13s of up to 5m 00s).
+148s The report was still processing (status: finalizing); waited 2m 24s for it to complete before reading its issues.
      Retrieved 7 accessibility issues (Total: 27)

Without this change the same call returns 14. New tests cover each source's status read (including a build found on the second page of its project's list, and an unreadable status), no wait when complete, waiting through finalizing (fake timers), the warning when processing outlasts the wait, no wait on later pages, and progress notifications. npm run build (lint, format, 1336 tests, tsc) passes.

API follow-up (optional): the Automated Tests build overview (GET /api/automated-tests/v1/builds/{id}) has no status, so the MCP reads it from the project's build list. get_build_summary already holds the TestHub object list_builds takes it from, so adding it there is one line and would make this a single call.

🤖 Generated with Claude Code

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

• Configuration used: Central YAML (base), Organization UI (inherited), Workspace UI (inherited)
• Review profile: ASSERTIVE
• Plan: Enterprise
• Run ID: 9ddebfbb-74e9-4218-8a27-b29cb0d7d1a5

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

_{Comment @coderabbitai help to get the list of available commands.}

[Aaryan430]

Superseded by #503, the same change raised from a branch in this repo instead of a fork, so the stacked PRs show only their own changes.