[FEATURE] Persist failing slash commands in the conversation log

Preflight Checklist

• [x] I have searched existing requests and this feature hasn't been requested yet
• [x] This is a single feature request (not multiple features)

Problem Statement

Running a valid slash command prints the command, eg:

❯ /clear
  ⎿  (no content)

but an invalid one will just return the output, eg running /asdf without a matching skill:

❯ Unknown skill: asdf

It's easy enough to infer, but feels like it breaks with the rest of the shell-like model.

Proposed Solution

I'd expect the input to persist like it does with successful commands:

❯ /asdf
  ⎿  Unknown skill: asdf

Alternative Solutions

_No response_

Priority

Low - Nice to have

Feature Category

Interactive mode (TUI)

Use Case Example

Example scenario:

1. I'm new to using CC and trying to set up a skill. I don't quite configure commands correctly, so running the slash command I expected to work results in ❯ Unknown skill: some-new-skill. I have a slightly hard time distinguishing what actually went wrong because my input disappeared, unlike most other interactions.

Additional Context

A side note, running into this issue and feeding it back into CC, Opus 4.6 said the skill registered by the plugin had a slightly different slash command, which also turned out to not be correct. So even Opus is a little bamboozled by the way skill slash command registration works here.