You'll noticeTier: how much it should matter to you
2Useful: my rating, 1 to 5
1Signal: worth watching, 1 to 5
PermissionsArea: what it touches
ImprovementsKind: in v2.1.235,
ImprovementsSection of the release
The tool-permission dialog now computes an "unrenderable" flag from its parameter entries, a failed tool-use render, or a withheld command, arguments, or URL. The always-allow option, "Yes, and don't ask again for...", is only offered when that flag is false, so if Claude Code can't display the whole command or its arguments, approval becomes one-time only.
A companion check also hides always-allow suggestions when the decision reason isn't classifier-approvable or the request is org-capped, and the rule label itself is dropped if it's too wide for the dialog.
How sure we are
One source agreesOne thing we can check says the same as this entry.
Anthropic's release notes agreeImproved permission dialogs: display text and "don't ask again" options now always match what a grant would cover, and "don't ask again" is…
