Track AMD SEV-SNP support gap on GCP

[h4x3rotab]

Context

While reviewing PR #743, we found a support-matrix gap: AMD SEV-SNP support should not be described or implied as available on GCP. The current GCP deployment path is Intel TDX.

Current state

• GCP docs and quickstart material are Intel TDX-focused.
• AMD SEV-SNP support is separate from the GCP path and depends on supported SEV-SNP hosts/images.
• Native TEE interface docs can mention AMD SEV-SNP generally, but provider-specific docs should make the GCP limitation explicit.

Tasks

• Audit README and docs for text that could imply AMD SEV-SNP is supported on GCP.
• Keep the support matrix explicit: GCP uses Intel TDX for now.
• Document where AMD SEV-SNP is supported once the v0.6.0 SEV-SNP image line is ready for users.
• Decide whether AMD SEV-SNP on GCP is a product goal. If yes, split follow-up work for provider/image/runtime prerequisites.

Related

• PR [codex] docs: add native TEE interface guide #743: native TEE interface docs and version availability.

[h4x3rotab]

Resolved. The support matrix in the README is already explicit that GCP uses Intel TDX (README.md:178), and I've added a one-sentence note that AMD SEV-SNP is new/experimental and that Intel TDX is the production path (README.md:180).

The strategic question this issue raised — "is AMD SEV-SNP on GCP a product goal?" — is answered in #713's platform-strategy section: GCP SNP, if pursued, is a separate provider-vTPM backend, not an extension of the bare-metal launch-measurement recompute path. It is not a current target. Closing the doc-truth part here; that future backend lives in #713.