SUBMISSION-STATUS — CGGMP21 (Lux Profile)

Honest framing. Tier B — production implementation, Lux-profile submission documentation in progress, formal submission gated per §3 below.

§1 Tier classification

Tier	Meaning	Status
A	Cut-ready submission package: spec consolidated, KAT manifest enforced, interop suites green, cut script verified	not yet
B	Implementation production-grade; submission-shape docs being assembled; gaps explicit; not deadline-bound	current
C	Implementation only; no submission scaffold	past state

Cross-suite comparison:

CGGMP21 is not a NIST MPTC primary candidate (NIST has not standardized threshold-ECDSA). The Lux profile targets:

Track	Form	Status
Lux-profile precompile spec	LP-4720	Final
Lux-profile precompile package (this dir)	7 docs (README/SPEC/SUBMISSION-STATUS/PROOF-CLAIMS/PARAMS/TEST-VECTORS/SECURITY)	Tier B in progress
Academic/IETF formal submission	NIST has no MPTC track for threshold-ECDSA; possible CFRG draft authorship is roadmap	not pursued
ACVP / CAVP for the underlying ECDSA	Tracks `luxfi/crypto/secp256k1` upstream — not separately validated here	n/a

The Lux production target is a Tier A submission tarball under scripts/cut-submission.sh bundling SPEC + ref impl + KAT + interop.

Status: KAT generator exists in tests; no Lux-specific KAT manifest yet.
Gate: stand up cmd/cmp_oracle/ (mirroring corona/cmd/corona_oracle_v2/); add regen-kats.sh --verify invariant.
Estimate: 2-3 weeks.

Status: tests cover Lux implementation; cross-validation against crypto/ecdsa (Go stdlib) + Bitcoin Core ECDSA verifier
- Ethereum geth ECDSA verifier is partial.
Gate: explicit cross-validation test exercising every signature against ≥3 third-party secp256k1 ECDSA verifiers.
Estimate: 1-2 weeks.

Status: pkg/paillier exists; biprime testing per CCS '21 Appendix C is implemented.
Gate: published audit of Paillier-modulus generation against CCS '21 §6.1 mandates (specifically: prime-quality, blum-prime property, biprime soundness).
Estimate: 1 week internal review + lab engagement for external.

Status: SPEC.md + LP-4720 + CCS '21 paper.
Gate: single spec/cmp-lux.tex consolidating these for reviewer convenience.
Estimate: weeks.

Status: benchmark tests exist (cmp_benchmark_test.go); no formal performance memo.
Gate: measured-vs-paper table (per-round latency, CPU cost, Paillier-op count) with reproducibility script.
Estimate: 1-2 weeks.

Status: round-N share verification implemented; explicit Lux blame-attribution flow on partition/equivocation is undocumented.
Gate: explicit IDENTIFIABLE-ABORT.md (or section in SPEC.md) with the Lux profile's blame rules.
Estimate: 1 week.

Status: no EasyCrypt theory, no Lean bridge, no Jasmin sources. Consistent with Corona's honest disclosure.
Gate: EC theory shell for the Lux profile.
Estimate: 8-16 weeks research + 12-16 weeks engineering. CGGMP21's UC framework + complex ZK subprotocols make this the most labor-intensive of the threshold primitives.

Status: no formal sign-off doc.
Gate: independent reviewer attests Lux profile correctness + CCS '21 conformance + Paillier handling + ZK subprotocol correctness.
Estimate: depends on reviewer engagement; multi-month.

Per the project's "no AI slop / no fake closure" rule, this package will NOT claim:

Mechanized refinement until §3.7 closes
Cryptographer sign-off until §3.8 closes
NIST MPTC Class N1 byte-equality framing — NIST has not standardized threshold-ECDSA; the analogous claim here is "byte-identical to single-party secp256k1 ECDSA" which IS made in SPEC.md §2.2.