You'll noticeTier: how much it should matter to you
2Useful: my rating, 1 to 5
3Signal: worth watching, 1 to 5
WorkflowsArea: what it touches
ImprovementsKind: in v2.1.300,
ImprovementsSection of the release
What
Workflow agents are helpers that Claude Code starts as part of a workflow run. When one starts, its prompt may now be a rewritten version, and a rewritten prompt goes through its own check. The model the agent uses is now chosen with the knowledge that it is running inside a workflow.
When a workflow is handed back to a session in dontAsk mode, a permission mode that never asks you for approval, Claude Code now adds this note: "This session's permission mode cannot ask for approval, so the run will not continue in this session unless that call is already allowed."
Why
This changes how workflow helpers pick their prompt and model. In dontAsk sessions the note explains why a handed-back run may stop.
The entry above is what we published on the day. These lines were added later, as Anthropic's own pages caught up, and they sit beside the original rather than replacing it.
Added sinceA small documentation edit on Run Claude Code programmatically touched a line naming dontAsk after this was published.To set a baseline for the whole session instead of listing individual tools, pass a [permission mode](/docs/en/permission-modes) with `--permission-mode`, such as `auto`, `dontAsk`, or `acceptEdits`:headlesssee the edit
Confirmed sinceAnthropic's documentation has since written up dontAsk, on Automate actions with hooks.* **`"deny"` in every permission mode**: a hook that returns `permissionDecision: "deny"` blocks the tool even in `bypassPermissions` mode or with `--dangerously-skip-permissions`. `PreToolUse` hooks fire before any permission-mode check, …hooks-guidesee the edit
How sure we are
One source agreesOne thing we can check says the same as this entry.
Anthropic's documentation agreesAnthropic's documentation has since written up dontAsk, on Automate actions with hooks.
