Seed Round: $0.25/FRAC — Listing at $1.00 (300% ROI) — Buy Now
Back to all standards
WAVE 1 STANDARD · v1.0 PUBLISHED · CC BY 4.0 · 2026-05-16

FQS-1

Fractal Quantum Shield Standard v1

Open standard for defense-in-depth post-quantum cryptography. Four components — QIE v2 + Cascade AEAD + Quantum Entropy Mixer + Drand adapter — all shipped, tested, and verified against real public randomness.

FQS-1 v1.0 — Four shipped components

Each component is independently specified, independently tested, and composes into the full FQS-1 stack.

Component 1

QIE v2 — Quantum Intelligent Encryption

✓ SHIPPED

AES-256-GCM-SIV + HMAC-SHA3-512 outer tag + HKDF-SHA3-256 key derivation + constant-time verification. MAC-then-decrypt. Zeroize-on-drop.

blockchain/crypto/src/qie_v2.rs· 15 tests· ~480 LOC
Component 2

Cascade AEAD — Triple defense-in-depth

✓ SHIPPED

AES-GCM-SIV → ChaCha20-Poly1305 → AES-GCM-SIV with HMAC outer tag. Two cipher families. Independent keys via HKDF. Per-layer domain separation.

blockchain/crypto/src/cascade_aead.rs· 17 tests· ~546 LOC
Component 3

Quantum Entropy Mixer

✓ SHIPPED

Combines N entropy sources via SHA3-512 → HKDF expansion. Defense-in-depth: attacker must compromise all sources. Graceful degradation on source failure.

blockchain/crypto/src/quantum_entropy.rs· 20 tests· ~487 LOC
Component 4

DrandSource — External verifiable entropy

✓ SHIPPED

Adapter to drand network (BLS-signed public randomness from 12+ validators). HTTP timeout + ring-buffer cache + graceful fallback. Verified against live drand round 6118315.

blockchain/node/src/drand_source.rs· 11 tests· ~310 LOC

The full FQS-1 stack composed

The components compose. The mixer feeds nonces with per-layer domain separation; the cascade encrypts using them. For an attacker to forge a ciphertext, they need to break two cipher families simultaneously AND compromise OS RNG plus drand network at the same time.

use fractal_crypto::{CascadeEngine, CascadeMasterKey, FractalEntropyMixer}; use fractal_node::drand_source::DrandSource; use std::sync::Arc; // 1. Mixer with OS RNG floor + drand external source let mixer = FractalEntropyMixer::new() .with_source(Arc::new(DrandSource::new())); // 2. Derive 3 nonces with per-layer domain separation let n1 = mixer.generate_nonce_12(b"fqs-cascade-l1-aes")?; let n2 = mixer.generate_nonce_12(b"fqs-cascade-l2-chacha")?; let n3 = mixer.generate_nonce_12(b"fqs-cascade-l3-aes")?; // 3. Encrypt with triple defense-in-depth let engine = CascadeEngine::new(CascadeMasterKey::generate()); let ct = engine.encrypt_with_nonces(plaintext, aad, &n1, &n2, &n3)?; let recovered = engine.decrypt(&ct, aad)?;

E2E verification: see fqs_stack_e2e_cascade_with_mixer_no_network + fqs_stack_e2e_cascade_with_drand_live in drand_source.rs

Why a Fractal Quantum Shield

Defense in depth, not single layer

TLS, NaCl, AWS KMS all use one cipher. FQS-1 uses two families plus an HMAC outer tag — breaking one is not enough.

Auditable external entropy

OS RNG is opaque. Drand network is publicly verifiable BLS-signed randomness from 12+ independent validators. FQS-1 mixes both.

Open spec, not vendor lock-in

Spec under CC BY 4.0, implementations under MIT. Anyone can implement, audit, critique, or fork. No proprietary primitives.

Honest caveats — what FQS-1 is NOT

  • ❌ NOT a replacement for NIST FIPS 203/204 (Kyber, Dilithium) — it complements them. KEM still needs Kyber.
  • ❌ NOT externally audited yet — Wave 3 (Q3 2026) includes Trail of Bits / Cure53 / NCC Group level audit.
  • ❌ NOT faster than single AES-GCM — Cascade is ~3× slower (the cost of defense-in-depth).
  • ❌ NOT a forward-secrecy protocol on its own — needs Kyber KEM per session for FS.
  • ❌ NOT protection against physical attacks (cold boot, side-channel hardware).

We publish what works. We publish what doesn't. Trust requires honesty about both.

Roadmap — extensions beyond v1.0

FQS-1 v1.0 is shipped. These extensions are tracked publicly for Waves 2 and 3.

Wave 2

HTTP API endpoints (Cascade + Mixer + Drand)

June 2026

Expose Cascade and entropy mixer via FractalPQC API: /v1/cascade/encrypt, /v1/cascade/decrypt, /v1/entropy/mixed, /v1/entropy/drand/status.

Wave 2

Multi-family PQC (Phase 2)

June 2026

Add SLH-DSA, HQC, Classic McEliece, CSIDH adapters. Diversification against single-family break.

Wave 2

ANU QRNG adapter

June 2026

Second external verifiable entropy source — Australian National University quantum random number generator. Makes mixer N≥3.

Wave 2

Threshold DKG + FROST (Phase 4)

July 2026

k-of-n threshold signing. Treasury and multisig without single point of failure.

Wave 3

Formal verification (Phase 6)

Q3 2026

Kani + Miri in CI. Mathematical proofs of constant-time properties. Fuzz testing 10K+ CPU hours.

Wave 3

External cryptographic audit (Phase 8)

Q3 2026

Trail of Bits / Cure53 / NCC Group level external audit of the published spec and reference implementations.

Wave 3

Post-Quantum Bridges (Phase 7)

Q3 2026

Migrate multi-chain bridge signatures from ECDSA to Dilithium. Future-proof against quantum computing arrival.

Wave 3

Real FHE integration

Q4 2026

TFHE / CKKS for private computation on encrypted data.

Source, spec, and critique

Read FQS-1 spec Reference implementations Submit critique

Spec license: CC BY 4.0 · Implementations license: MIT · External audit pending Wave 3 · Use crypto only if you understand it