Map SNP attestation to JSON/Rego types

The current rego policies are passed a couple of hardcoded fields (measurements, host_data), but should receive a fully parsed-out attestation that they can choose to interpret, and assert things like TCB versions for example.

I think the existing TypeScript mapping exposed by the CCF JS API is a good starting point for this, although I am not sure how Rego deals with ArrayBuffers: https://microsoft.github.io/CCF/main/js/ccf-app/interfaces/global.SnpAttestationResult.html

interface TcbVersion {
    boot_loader: number;
    microcode: number;
    snp: number;
    tee: number;
}

interface SnpAttestationResult {
    attestation: {
        author_key_digest: ArrayBuffer;
        chip_id: ArrayBuffer;
        committed_build: number;
        committed_major: number;
        committed_minor: number;
        committed_tcb: TcbVersion;
        cpuid_fam_id: number;
        cpuid_mod_id: number;
        cpuid_step: number;
        current_build: number;
        current_major: number;
        current_minor: number;
        family_id: ArrayBuffer;
        flags: {
            author_key_en: number;
            mask_chip_key: number;
            signing_key: number;
        };
        guest_svn: number;
        host_data: ArrayBuffer;
        id_key_digest: ArrayBuffer;
        image_id: ArrayBuffer;
        launch_tcb: TcbVersion;
        measurement: ArrayBuffer;
        platform_info: { smt_en: number; tsme_en: number };
        platform_version: TcbVersion;
        policy: {
            abi_major: number;
            abi_minor: number;
            debug: number;
            migrate_ma: number;
            single_socket: number;
            smt: number;
        };
        report_data: ArrayBuffer;
        report_id: ArrayBuffer;
        report_id_ma: ArrayBuffer;
        reported_tcb: TcbVersion;
        signature: { r: ArrayBuffer; s: ArrayBuffer };
        signature_algo: number;
        version: number;
        vmpl: number;
    };
    uvm_endorsements?: { did: string; feed: string; svn: string };
}

Items to resolve:

• What's the best way to expose binary buffers to rego-cpp now, and eventually?
• Should we line up UVM endorsements on SCITT CWT Claims, with the expectation that they will move to ESRP+CTS?
• Is it worth exposing the signature field at all? That will always be verified beforehand, there does not seem to be a use-case for that.