Public paste helpers should use real clipboard write plus real paste gesture

What happened

The first candidate public paste helpers for slate-browser were synthetic paste event tricks.

They looked small. They were not honest enough for a public API.

The real question was:

• can the browser write actual clipboard content
• can the test trigger an actual paste gesture
• does Slate handle that path correctly?

In Chromium on the Slate examples, yes.

What fixed it

The public paste helpers now use:

1. real clipboard write
   • navigator.clipboard.writeText(...)
   • or navigator.clipboard.write([ClipboardItem(...)])
2. real paste gesture
   • ControlOrMeta+V

That is a materially better public contract than fabricating a paste event and hoping the same path runs.

Why this matters

Clipboard architecture still stays clean:

• core/editor owns fragment meaning
• browser transport owns clipboard transport
• the public helper should prove the real transport path when it claims to test paste

Synthetic paste events are still fine for narrow legacy tests or experiments.

They are bad as the public default once a real browser path exists.

Reusable rule

If slate-browser/playwright exposes paste helpers, they should:

• write to the real browser clipboard
• trigger a real paste gesture
• avoid synthetic paste event hacks as the public default

If the real path stops working in a browser, that helper should be considered browser-limited there, not quietly downgraded back to fake paste.