fix: Improve error messages for missing build artifacts

When a test fails due to missing build artifacts, users previously saw
only a generic "An error occurred" message. This made it difficult to
understand what went wrong.

Changes:
- Add `error_message` property to Test model to expose the last error
- Update test detail template to show error message prominently in an
  alert box instead of hiding it in the progress table
- Add `_diagnose_missing_artifact()` helper that checks GitHub workflow
  status and provides specific error messages:
  - "Build still in progress" - when the GitHub Actions build hasn't
    completed yet (common for Windows builds which take ~40 minutes)
  - "Build failed" - when the workflow failed
  - "Artifact not found" - when build succeeded but artifact missing
    (possibly expired)
  - "No workflow run found" - when the workflow hasn't started

This helps users understand why their test failed and what action to
take (e.g., wait for build to complete, check GitHub Actions logs).

Fixes the issue where test 7139 showed a cryptic error because the
Windows build was still in progress when the test tried to fetch the
artifact.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

Author: cfsmp3

mod_ci/controllers.py

@@ -304,6 +304,57 @@ def mark_test_failed(db, test, repository, message: str) -> None:
         log.error(f"Failed to mark test {test.id} as failed: {e}")
 
 
+def _diagnose_missing_artifact(repository, commit_sha: str, expected_workflow: str, log) -> str:
+    """
+    Diagnose why an artifact was not found for a commit.
+
+    Checks the workflow run status to provide a more helpful error message.
+
+    :param repository: GitHub repository object
+    :param commit_sha: The commit SHA to check
+    :param expected_workflow: The expected workflow name (e.g., "Build CCExtractor on Windows")
+    :param log: Logger instance
+    :return: A descriptive error message
+    """
+    from collections import defaultdict
+
+    try:
+        # Build workflow name lookup
+        workflow = defaultdict(lambda: None)
+        for active_workflow in repository.get_workflows():
+            workflow[active_workflow.id] = active_workflow.name
+
+        # Check workflow runs for this commit
+        workflow_found = False
+        for workflow_run in repository.get_workflow_runs(head_sha=commit_sha):
+            workflow_run_name = workflow[workflow_run.workflow_id]
+            if workflow_run_name != expected_workflow:
+                continue
+
+            workflow_found = True
+            if workflow_run.status != "completed":
+                return (f"Build still in progress: '{expected_workflow}' is {workflow_run.status}. "
+                        f"Please wait for the build to complete and retry.")
+            elif workflow_run.conclusion != "success":
+                return (f"Build failed: '{expected_workflow}' finished with conclusion '{workflow_run.conclusion}'. "
+                        f"Check the GitHub Actions logs for details.")
+            else:
+                # Build succeeded but artifact not found - may have expired
+                return (f"Artifact not found: '{expected_workflow}' completed successfully, "
+                        f"but no artifact was found. The artifact may have expired (GitHub deletes "
+                        f"artifacts after a retention period) or was not uploaded properly.")
+
+        if not workflow_found:
+            return (f"No workflow run found: '{expected_workflow}' has not run for commit {commit_sha[:7]}. "
+                    f"This may indicate the workflow was not triggered or is queued.")
+
+    except Exception as e:
+        log.warning(f"Failed to diagnose missing artifact: {e}")
+        return f"No build artifact found for this commit (diagnostic check failed: {e})"
+
+    return "No build artifact found for this commit"
+
+
 def start_test(compute, app, db, repository: Repository.Repository, test, bot_token) -> None:
     """
     Start a VM instance and run the tests.
@@ -439,8 +490,10 @@ def start_test(compute, app, db, repository: Repository.Repository, test, bot_to
     artifacts = repository.get_artifacts()
     if test.platform == TestPlatform.linux:
         artifact_name = Artifact_names.linux
+        workflow_name = Workflow_builds.LINUX
     else:
         artifact_name = Artifact_names.windows
+        workflow_name = Workflow_builds.WINDOWS
     for index, artifact in enumerate(artifacts):
         if artifact.name == artifact_name and artifact.workflow_run.head_sha == test.commit:
             artifact_url = artifact.archive_download_url
@@ -474,8 +527,10 @@ def start_test(compute, app, db, repository: Repository.Repository, test, bot_to
             break
 
     if not artifact_saved:
-        log.critical("Could not find an artifact for this commit")
-        mark_test_failed(db, test, repository, "No build artifact found for this commit")
+        # Try to diagnose why no artifact was found by checking workflow run status
+        error_detail = _diagnose_missing_artifact(repository, test.commit, workflow_name, log)
+        log.critical(f"Could not find artifact for commit {test.commit}: {error_detail}")
+        mark_test_failed(db, test, repository, error_detail)
         return
 
     zone = config.get('ZONE', '')

mod_test/models.py

@@ -183,6 +183,18 @@ def failed(self):
             return self.progress[-1].status == TestStatus.canceled
         return False
 
+    @property
+    def error_message(self):
+        """
+        Get the error message if the test failed.
+
+        :return: The error message from the last progress entry, or None if not failed
+        :rtype: str or None
+        """
+        if self.failed and len(self.progress) > 0:
+            return self.progress[-1].message
+        return None
+
     @property
     def github_link(self):
         """

templates/test/by_id.html

@@ -83,7 +83,12 @@ <h1>Test progress for {{ title }}</h1>
                 {% if test.finished %}
                     {% if test.failed %}
                         <h2>An error occurred</h2>
-                        <p>Something went wrong while the tests where running. You should be able to see in the bar above during which stage it went wrong exactly.</p>
+                        {% if test.error_message %}
+                        <div class="callout alert">
+                            <strong>Error:</strong> {{ test.error_message }}
+                        </div>
+                        {% endif %}
+                        <p>Something went wrong while the tests were running. You can see the progress timeline above for more details.</p>
                         <p>If you have build errors, you can access the above log file with build errors.</p>
                     {% else %}
                         <h2>Test results</h2>