fix: update MXFP8 kernels for NKI 0.3.0

[mrkcath-aws]

Issue #, if available:

N/A

Description of changes:

• Remove stale dtype kwarg from tensor_copy
• Remove deprecated platform_target from @nki.jit
• Fix arg order: dst before src (0.3.0 convention)
• Add buffer assertions for kernel 3
• Add 4th kernel: packed scale variant
• Add doc markers for nki_example directive

Synced from private-aws-neuron-sdk-staging. All 4 kernels verified on trn3.

Testing:

Please see detailed unit test requirements in the CONTRIBUTING.md

• [ N/A ] The change is covered by numeric check using nki.baremetal
• [ N/A ] The change is covered by performance benchmark test using nki.benchmark
• [ x ] The change is covered by end-to-end integration test

Pull Request Checklist

• [ x ] I have filled in all the required field in the template
• [ x ] I have tested locally that all the tests pass
• [ x ] By submitting this pull request, I confirm that my contribution is made under the terms of the MIT-0 license.