fix(runner): The runner help is wrong about screens, --name, auto-launch and exit codes

[Alex Pezzati (alex-pezzati)]

Overview of Problem

As an agent driving a QA Wolf runner from the CLI, I want the runner help to describe what runners actually do, so that I don't run flows I don't need or misread what a command did.

Context

• The help says a runner has no screen until its first run, so agents run a flow before acting. A playwright runner's screen now starts at launch.
• The help says a different --name on a running id is refused, but it attaches. It also leaves actions out of the commands that launch a runner, and doesn't list exit code 7.

Overview of Changes

• Only text changes, in the help, error messages, the skill file and code comments. No command behaves differently.
• The runner help says a playwright runner has a screen as soon as it launches, and that any other runner still needs a run first.
• The launch help says a different --name attaches to the running runner instead of being refused.
• The workflow guide lists actions among the commands that launch a runner, and explains exit code 7.
• The errors screenshot, inspect and highlight print when no runner is named say the same thing.

Fixes AGE-5935

Testing

bun run typecheck
bun run lint
bun run format:check
bun run knip
bun run test

Checklist

• Changes follow the code style of this project
• Self-review completed
• Tests added/updated (or not applicable)
• No breaking changes (or described below)

[Alex Pezzati (alex-pezzati)]

kn-content These are the CLI help fixes you flagged in your comment on the docs PR. They cover --name attaching, a browser runner's screen starting at launch, actions launching a runner, and exit code 7. Nothing needed from you, just keeping you in the loop.

[Felipe Augusto (felipe-augusto)]

Alex Pezzati (@alex-pezzati) If you want a new release of the cli you need to bump the version