You'll noticeTier: how much it should matter to you
2Useful: my rating, 1 to 5
2Signal: worth watching, 1 to 5
AgentsArea: what it touches
ImprovementsKind: in v2.1.318,
ImprovementsSection of the release
What
An agent is a helper that Claude starts to work on part of a task. When Anthropic's safeguards stop one, Claude Code now shows a message that begins "Safeguards stopped this agent." The wording differs depending on whether automatic model switching is on.
When the reason is cybersecurity, the message suggests writing a clearer prompt if the task is legitimate. It can also mention that Defense Access in the Cyber Verification Program reduces these interruptions.
Why
If your security work keeps being stopped, the message now tells you what you can do about it.
