Where the workflow shifted
`ai-agent-workflow` has visibility but weak clicks; it needs more concrete permission and rollback language.
Define what can run automatically, what needs confirmation, how failures roll back, and who owns final review.
Tool names are not outcomes
The signal matters when it changes how a team ships, reviews, or recovers work, not when it only names another tool.
Check permissions and failure
- Align `ai-agent-workflow` entry text with permission, logging, test, and rollback facts
- Keep the test narrow: one low-risk task or tool entry before connecting permissions, logs, failure handling, and human takeover to production
What still needs proof
Missing confirmation and rollback facts make agent workflows look like risky black boxes. Keep the original source open so the announcement, the evidence, and this site's interpretation stay separate.