[DOCS] Agent view docs omit bypass-mode disclaimer and per-spawn bypass behavior for `claude agents --dangerously-skip-permissions`

Documentation Type

Missing documentation (feature not documented)

Documentation Location

https://code.claude.com/docs/en/agent-view

Section/Topic

The "Permission mode, model, and effort" subsection that documents claude agents permission-mode flags.

Current Documentation

The page currently documents the flag as a shorthand:

"claude agents also accepts --dangerously-skip-permissions as shorthand for --permission-mode bypassPermissions, and --allow-dangerously-skip-permissions to make bypassPermissions available in each dispatched session's Shift+Tab cycle without starting in that mode. Both match the top-level CLI flags."

"Using bypassPermissions or auto is refused until you have accepted that mode by running claude with it once interactively, since those modes let a session you aren't watching act without approval. The same applies whether you pass the mode to claude agents or to claude --bg --permission-mode."

The page does not describe what happens when the mode has not been accepted yet (a user types claude agents --dangerously-skip-permissions cold) or what the supervisor passes through to sessions the dispatcher spawns internally.

What's Wrong or Missing?

Before v2.1.196, passing --dangerously-skip-permissions to claude agents without having accepted bypass mode interactively first silently fell back to auto mode for dispatched sessions. v2.1.196 fixes this so the dispatcher shows the standard bypass disclaimer and applies bypassPermissions to the spawned agent sessions itself. Neither behavior — the disclaimer path nor the per-spawn propagation — is documented on the agent-view page. Users relying on --dangerously-skip-permissions from a CI runner or script have no way to confirm the bypass actually applied.

Suggested Improvement

Add a short note directly under the existing --dangerously-skip-permissions paragraph that documents the v2.1.196 behavior:

claude agents --dangerously-skip-permissions uses the same bypass warning and acceptance flow as the top-level --dangerously-skip-permissions flag. After the warning is accepted, the bypass mode applies to sessions the dispatcher spawns. It should not silently fall back to auto when the user explicitly requested bypass mode.

Update the surrounding "Permission mode, model, and effort" prose to make the dispatcher inheritance explicit: sessions started from an agent-view process launched with bypass mode should inherit that requested mode after the required warning is accepted.

Impact

High - Prevents users from using a feature

Additional Context

Affected Pages:
| Page | Context |
|------|---------|
| https://code.claude.com/docs/en/agent-view | Primary location for claude agents --dangerously-skip-permissions behavior |
| https://code.claude.com/docs/en/permission-modes | Cross-reference for --dangerously-skip-permissions acceptance flow and disclaimer |
| https://code.claude.com/docs/en/cli-reference | Cross-reference for the top-level flag |

Total scope: 3 pages affected.

v2.1.196 makes claude agents --dangerously-skip-permissions show the bypass disclaimer and propagate bypass mode to spawned agent sessions. The agent-view page currently only documents the shorthand equivalence to bypassPermissions; the disclaimer path and per-spawn propagation are new and need explicit coverage.