Use direct language for error messages

The bot uses indirect phrasing for error messages, which takes more effort to understand and introduces doubt:

`You seem to already be following ${url}`
`It seems you were not following ${blog}`

I'd recommend using more direct phrasing like this:

`You are already following ${url}`
`You were not following ${blog}`

Search the bot source for the two indirect error-message phrases shown in the issue and inspect the follow and unfollow paths that use them. Replace those messages with the direct wording proposed, then verify both message cases read naturally.

Written by the indexing model from the issue text.