We've seen an increased failure rate with Internal Errors.
I had my agent run an investigation and the results are below.
We have filed an incident with our infrastructure provider.
Confirmed. There is a broad dynamic-scan incident, but the evidence does not yet prove the GitHub App caused it.
• GitHub App production cutover: August 10, 20:10 UTC — #263724
• 231 PRs currently carry Internal-Error-Dynamic-Scan .
• In a measured August 12 window, 69 of 209 new PRs—33%—were affected. This is conservative because later validations were still running.
• The failures span many packages and contributors, not just the reporter.
• The reporter’s post-cutover series had 3 of 10 PRs affected, including both 1.32 packages:
• #416226
• #416227
• #415912
• Unrelated examples show the same signature:
• #416070
• #416388
The App exposes the validation stages as GitHub checks; dynamic scanning itself predates the migration, and similar infrastructure errors were already tracked at #411157. My conclusion is confirmed high failure rate and likely dynamic-scan infrastructure/capacity failure coincident with the cutover—not confirmed App regression.
We've seen an increased failure rate with Internal Errors.
I had my agent run an investigation and the results are below.
We have filed an incident with our infrastructure provider.
Confirmed. There is a broad dynamic-scan incident, but the evidence does not yet prove the GitHub App caused it.
• GitHub App production cutover: August 10, 20:10 UTC — #263724
• 231 PRs currently carry Internal-Error-Dynamic-Scan .
• In a measured August 12 window, 69 of 209 new PRs—33%—were affected. This is conservative because later validations were still running.
• The failures span many packages and contributors, not just the reporter.
• The reporter’s post-cutover series had 3 of 10 PRs affected, including both 1.32 packages:
• #416226
• #416227
• #415912
• Unrelated examples show the same signature:
• #416070
• #416388
The App exposes the validation stages as GitHub checks; dynamic scanning itself predates the migration, and similar infrastructure errors were already tracked at #411157. My conclusion is confirmed high failure rate and likely dynamic-scan infrastructure/capacity failure coincident with the cutover—not confirmed App regression.