[DOCS] Agent view docs omit no-errno spawn failure diagnostics for background dispatches

Documentation Type

Missing documentation (feature not documented)

Documentation Location

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

Section/Topic

Dispatch new agents, failed-session diagnostics, and troubleshooting for background dispatches that fail before the session starts

Current Documentation

The agent view page currently says:

Type a prompt in the input at the bottom of agent view and press Enter to start a new background session.

It later defines the failed state only as:

| Failed | Red | The task ended with an error |

And the shell command table says:

| claude logs <id> | Print the session's recent output | | claude daemon status | Print the supervisor's state, version, socket directory, and worker count |

The troubleshooting section also says:

If attaching, peeking, or claude logs reports that the background service did not respond, the supervisor process has likely stalled. Stop it and let the next claude agents start a fresh one.

What's Wrong or Missing?

The docs cover how to dispatch a background session, how to recognize a generic failed row, and how to inspect recent output, but they do not document the spawn-failure diagnostics added in v2.1.162.

The changelog for v2.1.162 says:

Background dispatch spawn failures now report the error class name when no errno is available

That leaves two missing pieces in the live docs:

A. Failed background dispatches have no documented diagnostic path

The current page tells readers how to start a session and how to inspect logs, but it does not explain what kind of failure details to expect when the dispatch itself fails before the session can start normally.

B. The class-name fallback is undocumented

When an OS-style errno is unavailable, Claude Code now reports the error class name instead. The docs do not mention that fallback, so users who see a class name instead of something like ENOENT have no documentation explaining that this is expected diagnostic output.

Without that note, immediate dispatch failures remain harder to interpret than they need to be, even though the product now surfaces a more useful identifier.

Suggested Improvement

Add a short troubleshooting note to the agent view page anywhere it discusses failed rows, claude logs <id>, or supervisor diagnostics.

For example:

If a background dispatch fails before the session fully starts, inspect the row with claude logs <id> or check the supervisor diagnostics. Spawn failures usually include an errno-style code such as ENOENT. When no errno is available, Claude Code reports the error class name instead so you still have a stable identifier for troubleshooting.

If the team prefers to keep the detailed explanation out of the main walkthrough, a short troubleshooting subsection for "dispatch failed before start" would still close the gap.

Impact

Medium - Makes feature difficult to understand

Additional Context

Affected Pages:

| Page | Line(s) | Context |
|------|---------|---------|
| https://code.claude.com/docs/en/agent-view | 109-115, 243-259, 448-489, 521-529 | Covers failed rows, dispatch flow, claude logs, claude daemon status, and adjacent troubleshooting, but not how spawn failures are identified when dispatch fails before start |
| https://code.claude.com/docs/en/cli-reference | 27-32 | Lists claude agents, claude logs, and claude daemon status as diagnostics commands without explaining the spawn-failure identifiers those commands may surface |

Total scope: 2 pages affected

Version context: v2.1.162 changelog: background dispatch spawn failures now report the error class name when no errno is available.