Saorsa Post-Quantum Cryptography Library

A comprehensive, production-ready Post-Quantum Cryptography library providing a complete quantum-secure cryptographic suite. Implements NIST FIPS 203, 204, and 205 standardized algorithms for asymmetric cryptography, plus comprehensive cryptographic primitives including BLAKE3, SHA3, HMAC, HKDF, AES-256-GCM, and ChaCha20-Poly1305. This library provides high-performance, thoroughly tested implementations with a clean, safe API, all validated against official NIST ACVP, RFC, and specification test vectors.

🎯 Features

• Complete Quantum-Secure Suite: Both asymmetric (PQC) and symmetric encryption with comprehensive cryptographic primitives
• FIPS 140-3 Compliant RNG: ChaCha20-based DRBG with continuous health monitoring per NIST SP 800-90A/B
• FIPS-Certified Implementations: Uses NIST FIPS-certified crates for ML-KEM, ML-DSA, and SLH-DSA
• Extensive Cryptographic Library: BLAKE3, SHA3, HMAC, HKDF, AES-256-GCM, ChaCha20-Poly1305, and HPKE
• Official Test Vector Validation: All algorithms validated against NIST ACVP, RFC, and specification test vectors
• Comprehensive API: Simple, user-friendly interfaces with FIPS-compliant RNG management
• High Performance: Optimized implementations with SIMD support where available
• Memory Safety: Automatic zeroization of sensitive data
• Type Safety: Strongly typed wrappers prevent misuse
• No Unsafe Code: Pure Rust implementations in the API layer
• Deterministic Testing: Support for reproducible key generation from seeds

📦 Installation

[dependencies]
saorsa-pqc = "0.4"

🔐 Supported Algorithms

🔒 Cryptographic Primitives (All Quantum-Resistant)

Hash Functions

• BLAKE3: Modern cryptographic hash with tree hashing

  • ✅ High Performance: Faster than SHA2/SHA3 with parallelization
  • ✅ 256-bit Output: Configurable output length (XOF capability)
  • ✅ Test Vectors: Validated against official BLAKE3 specification vectors
  • ✅ Use Cases: General hashing, key derivation, checksums

• SHA3-256/SHA3-512: NIST FIPS 202 Keccak-based hash functions

  • ✅ NIST Standard: FIPS 202 compliant implementation
  • ✅ Quantum Resistance: Based on different mathematical foundation than SHA2
  • ✅ Test Vectors: Validated against NIST FIPS 202 official test vectors
  • ✅ Use Cases: Digital signatures, certificates, blockchain applications

Key Derivation Functions (KDF)

• HKDF-SHA3-256/HKDF-SHA3-512: Extract-and-expand key derivation
  • ✅ RFC 5869 Based: Adapted for SHA3 hash functions
  • ✅ Secure Key Derivation: Extract entropy then expand to desired length
  • ✅ Test Vectors: Validated against RFC 5869 methodology
  • ✅ Use Cases: Deriving encryption keys from shared secrets

Message Authentication Codes (MAC)

• HMAC-SHA3-256/HMAC-SHA3-512: Hash-based message authentication
  • ✅ Constant-Time Verification: Resistant to timing attacks
  • ✅ NIST CAVS Tested: Validated against NIST test methodology
  • ✅ Flexible Key Sizes: Accepts arbitrary key lengths
  • ✅ Use Cases: Message integrity, authentication tokens

Authenticated Encryption (AEAD)

• AES-256-GCM: Hardware-accelerated authenticated encryption

  • ✅ Hardware Support: AES-NI acceleration on modern CPUs
  • ✅ 256-bit Security: Quantum-resistant key size
  • ✅ NIST CAVP Tested: Validated against NIST SP 800-38D test vectors
  • ✅ Use Cases: High-speed data encryption, VPN tunnels

• ChaCha20-Poly1305: Software-optimized authenticated encryption

  • ✅ Constant-Time: Resistant to side-channel attacks
  • ✅ 256-bit Security: Full 256-bit key strength
  • ✅ IETF Standard: RFC 8439 compliant
  • ✅ Test Vectors: Validated against RFC 8439 official test vectors
  • ✅ Use Cases: Mobile devices, embedded systems, general encryption

Hybrid Public Key Encryption (HPKE)

• HPKE with ML-KEM: RFC 9180 hybrid encryption bound to post-quantum KEMs
  • ✅ Post-Quantum: Combines ML-KEM with symmetric primitives
  • ✅ Multiple Modes: Base mode and PSK (pre-shared key) mode
  • ✅ Flexible Configuration: Choose KEM (ML-KEM variant), KDF, and AEAD
  • ✅ Test Vectors: Custom test vectors for ML-KEM combinations
  • ✅ Use Cases: End-to-end encryption, secure messaging, hybrid cryptosystems

ML-KEM (FIPS 203) - Key Encapsulation

• ML-KEM-512: NIST Level 1 (128-bit security)
• ML-KEM-768: NIST Level 3 (192-bit security)
• ML-KEM-1024: NIST Level 5 (256-bit security)

ML-DSA (FIPS 204) - Digital Signatures

• ML-DSA-44: NIST Level 2 (~128-bit security)
• ML-DSA-65: NIST Level 3 (~192-bit security)
• ML-DSA-87: NIST Level 5 (~256-bit security)

SLH-DSA (FIPS 205) - Stateless Hash-Based Signatures

12 variants covering all combinations of:

• Hash functions: SHA2, SHAKE
• Security levels: 128, 192, 256 bits
• Trade-offs: Small signatures (s) vs Fast signing (f)

💻 Quick Start

NEW: Trait-Based API (v0.3.11+)

The library now provides a trait-based abstraction layer for PQC algorithms, enabling easier integration and algorithm agility:

use saorsa_pqc::pqc::{Kem, Sig, MlKem768Trait, MlDsa65Trait, ConstantTimeCompare};

// Key Encapsulation with trait API
let (kem_pk, kem_sk) = MlKem768Trait::keypair();
let (shared_secret, ciphertext) = MlKem768Trait::encap(&kem_pk);
let recovered = MlKem768Trait::decap(&kem_sk, &ciphertext)?;
assert!(shared_secret.ct_eq(&recovered)); // Constant-time comparison

// Digital Signatures with trait API
let (sig_pk, sig_sk) = MlDsa65Trait::keypair();
let message = b"Quantum-resistant message";
let signature = MlDsa65Trait::sign(&sig_sk, message);
assert!(MlDsa65Trait::verify(&sig_pk, message, &signature));

// BLAKE3 helpers for secure key derivation
use saorsa_pqc::pqc::blake3_helpers;
let derived_key = blake3_helpers::derive_key("app-context", shared_secret.as_ref());

Key Features of Trait API:

• Zero-copy buffers: Efficient memory usage without unnecessary allocations
• Automatic zeroization: Secret keys are wiped from memory when dropped
• Constant-time operations: Protection against timing attacks
• Generic programming: Write code that works with any KEM or signature algorithm

Quantum-Secure Symmetric Encryption (ChaCha20-Poly1305)

use saorsa_pqc::api::ChaCha20Poly1305;
use saorsa_pqc::api::symmetric::{generate_key, generate_nonce};

// Generate a random 256-bit key (quantum-secure)
let key = generate_key();
let cipher = ChaCha20Poly1305::new(&key);

// Encrypt data with authenticated encryption
let nonce = generate_nonce(); // 96-bit nonce
let plaintext = b"Secret quantum-secure message";
let aad = b"Additional authenticated data";

// Encrypt with associated data (AEAD)
let ciphertext = cipher.encrypt_with_aad(&nonce, plaintext, aad)?;

// Decrypt and verify authenticity
let decrypted = cipher.decrypt_with_aad(&nonce, &ciphertext, aad)?;

assert_eq!(&decrypted[..], plaintext);

// Simple encryption without AAD
let ciphertext2 = cipher.encrypt(&nonce, plaintext)?;
let decrypted2 = cipher.decrypt(&nonce, &ciphertext2)?;
assert_eq!(&decrypted2[..], plaintext);

Key Encapsulation (ML-KEM)

use saorsa_pqc::api::{ml_kem_768, MlKemPublicKey, MlKemSecretKey};

// Generate keypair (RNG handled internally)
let kem = ml_kem_768();
let (public_key, secret_key) = kem.generate_keypair()?;

// Encapsulate - creates shared secret and ciphertext
let (shared_secret, ciphertext) = kem.encapsulate(&public_key)?;

// Decapsulate - recovers shared secret from ciphertext
let recovered_secret = kem.decapsulate(&secret_key, &ciphertext)?;

assert_eq!(shared_secret.to_bytes(), recovered_secret.to_bytes());

Digital Signatures (ML-DSA)

use saorsa_pqc::api::{ml_dsa_65, MlDsaPublicKey, MlDsaSecretKey};

// Generate keypair
let dsa = ml_dsa_65();
let (public_key, secret_key) = dsa.generate_keypair()?;

// Sign message
let message = b"Authenticate this message";
let signature = dsa.sign(&secret_key, message)?;

// Verify signature
let is_valid = dsa.verify(&public_key, message, &signature)?;
assert!(is_valid);

Stateless Signatures (SLH-DSA)

use saorsa_pqc::api::{slh_dsa_sha2_128s, SlhDsaPublicKey, SlhDsaSecretKey};

// Generate keypair (note: SLH-DSA keygen is slow)
let slh = slh_dsa_sha2_128s();
let (public_key, secret_key) = slh.generate_keypair()?;

// Sign and verify
let message = b"Quantum-resistant message";
let signature = slh.sign(&secret_key, message)?;
let is_valid = slh.verify(&public_key, message, &signature)?;
assert!(is_valid);

🧪 Testing & Validation

This library has been extensively tested against official test vectors from multiple authoritative sources:

Comprehensive Test Vector Validation

Post-Quantum Algorithms (NIST ACVP)

• Official NIST ACVP Vectors: github.com/usnistgov/ACVP-Server
  • ✅ ML-KEM: Keygen, Encapsulation/Decapsulation
  • ✅ ML-DSA: Keygen, Signature Generation/Verification
  • ✅ SLH-DSA: Comprehensive test vectors for all 12 variants

Cryptographic Primitives (Official Standards)

• ✅ BLAKE3: Official specification test vectors from BLAKE3 team
• ✅ SHA3-256/512: NIST FIPS 202 test vectors for empty input, "abc", and multi-million character tests
• ✅ AES-256-GCM: NIST CAVP test vectors from SP 800-38D with various key, IV, and AAD combinations
• ✅ HKDF-SHA3: Test vectors adapted from RFC 5869 methodology for SHA3 variants
• ✅ HMAC-SHA3: Test vectors derived from NIST CAVS testing methodology
• ✅ ChaCha20-Poly1305: RFC 8439 official test vectors
• ✅ HPKE: RFC 9180 methodology adapted for ML-KEM combinations

Additional Test Sources

• C2SP/CCTV Test Vectors: github.com/C2SP/CCTV
  • Intermediate values for debugging
  • Invalid input testing (modulus vectors)
  • Edge case testing (unlucky NTT sampling)

Running Tests

# Run all tests
cargo test --all-features

# Run specific algorithm tests
cargo test --test nist_official_vectors
cargo test --test extended_crypto_vectors

# Run with release optimizations (faster)
cargo test --release

Test Coverage

Post-Quantum Algorithm Coverage

• ✅ Key generation deterministic tests
• ✅ Encapsulation/Decapsulation correctness
• ✅ Signature generation and verification
• ✅ Wrong message/ciphertext rejection
• ✅ Serialization round-trips
• ✅ Context handling (ML-DSA)
• ✅ Parameter validation
• ✅ Cross-implementation compatibility

Cryptographic Primitive Coverage

• ✅ Hash Functions: BLAKE3 (empty, single byte, multi-part, million character), SHA3-256/512 (NIST FIPS 202)
• ✅ Key Derivation: HKDF-SHA3-256/512 deterministic output, salt handling, different context values
• ✅ Message Authentication: HMAC-SHA3-256/512 with various key sizes, constant-time verification
• ✅ AEAD Encryption: AES-256-GCM and ChaCha20-Poly1305 with AAD, authentication failure detection
• ✅ HPKE: All ML-KEM variants with different KDF/AEAD combinations, wrong key rejection
• ✅ Security Properties: Memory zeroization, constant-time operations, authentication tag verification
• ✅ Error Handling: Invalid input sizes, wrong authentication tags, corrupted data

📊 Performance Benchmarks

Run comprehensive benchmarks:

cargo bench --bench comprehensive_benchmarks

Benchmark Results (M1 Pro)

Algorithm	Operation	Time	Throughput
ML-KEM-768	KeyGen	~50 μs	-
ML-KEM-768	Encapsulate	~55 μs	-
ML-KEM-768	Decapsulate	~65 μs	-
ML-DSA-65	KeyGen	~120 μs	-
ML-DSA-65	Sign	~350 μs	-
ML-DSA-65	Verify	~130 μs	-
SLH-DSA-SHA2-128f	KeyGen	~3 ms	-
SLH-DSA-SHA2-128f	Sign	~90 ms	-
SLH-DSA-SHA2-128f	Verify	~4 ms	-
ChaCha20-Poly1305	Encrypt (1KB)	~0.8 μs	1.25 GB/s
ChaCha20-Poly1305	Encrypt (64KB)	~12 μs	5.3 GB/s
ChaCha20-Poly1305	Decrypt (64KB)	~12 μs	5.3 GB/s
AES-256-GCM	Encrypt (1KB)	~0.6 μs	1.67 GB/s
AES-256-GCM	Encrypt (64KB)	~8 μs	8.0 GB/s
BLAKE3	Hash (1KB)	~0.4 μs	2.5 GB/s
SHA3-256	Hash (1KB)	~1.2 μs	833 MB/s
HMAC-SHA3-256	MAC (1KB)	~1.3 μs	769 MB/s

Note: Performance varies by hardware. AES-GCM benefits from AES-NI acceleration. ChaCha20-Poly1305 and BLAKE3 benefit from SIMD acceleration (AVX2/NEON).

🔒 Security Considerations

Quantum Security

• Symmetric Algorithms: All symmetric algorithms (AES-256-GCM, ChaCha20-Poly1305) provide quantum resistance with 256-bit keys, offering 128-bit quantum security against Grover's algorithm
• Hash Functions: BLAKE3 and SHA3 maintain security against quantum attacks as they're based on different mathematical foundations
• Post-Quantum Asymmetric: ML-KEM, ML-DSA, and SLH-DSA are specifically designed to resist both classical and quantum attacks
• Complete Protection: Use ML-KEM for key exchange, then derive symmetric keys for AES-256-GCM or ChaCha20-Poly1305 encryption
• Algorithm Selection Guide:
  • Performance Priority: BLAKE3 (hashing), AES-256-GCM (encryption if AES-NI available)
  • Security Priority: SHA3 (standardized), ChaCha20-Poly1305 (constant-time)
  • Compatibility: SHA3 and AES-256-GCM (NIST standards)
  • Embedded/Mobile: BLAKE3 and ChaCha20-Poly1305 (software-optimized)

Implementation Security

1. Memory Safety: All sensitive data is automatically zeroized on drop
2. Constant Time: Critical operations verified constant-time via DudeCT statistical analysis
3. FIPS 140-3 RNG: ChaCha20-based DRBG with continuous health monitoring (SP 800-90A/B compliant)
4. No Key Reuse: Fresh randomness for each operation requiring it
5. Input Validation: All inputs validated before cryptographic operations
6. AEAD Protection: Both AES-256-GCM and ChaCha20-Poly1305 provide confidentiality and authenticity
7. Algorithm Diversity: Multiple implementations allow for algorithm agility and risk mitigation
8. Test Vector Compliance: All implementations validated against official standards
9. Formal CT Verification: DudeCT benchmarks in CI ensure timing attack resistance

Side-Channel Protection (v0.4.0+)

The library implements comprehensive side-channel protection with formal verification:

Protection	Implementation	Verification
Constant-time comparison	ct_eq(), ct_array_eq()	DudeCT verified (max_t < 5)
Constant-time selection	ct_select(), ct_assign()	DudeCT verified
Constant-time copy	ct_copy_bytes()	DudeCT verified
AAD hash verification	ct_eq() in AEAD	Prevents timing oracle
Memory clearing	zeroize crate	Compiler-resistant zeroing

DudeCT Integration: Every PR is automatically tested for timing leaks:

• Statistical analysis compares timing distributions between input classes
• |max_t| > 5 indicates non-constant-time behavior (CI fails)
• Extended weekly analysis runs 5-minute tests per primitive

# Run constant-time verification locally
cargo bench --bench ct_verification

FIPS 140-3 Compliance

The library implements a FIPS 140-3 compliant random number generator:

• DRBG Mechanism: ChaCha20-based deterministic random bit generator (approved for FIPS 140-3)
• Continuous Health Monitoring: Implements Repetition Count Test (RCT) and Adaptive Proportion Test (APT) per NIST SP 800-90B
• Entropy Source Validation: Startup health tests and continuous monitoring of OS entropy
• Automatic Reseeding: Reseeds after 1MB of output to ensure prediction resistance and backtracking resistance
• Security Strengths: Supports 128-bit, 192-bit, and 256-bit security levels
• Compliance: Meets requirements of NIST SP 800-90A (DRBG) and SP 800-90B (Entropy Sources)

use saorsa_pqc::pqc::{FipsRng, SecurityStrength};

// Create FIPS-compliant RNG for 256-bit security
let mut rng = FipsRng::new(SecurityStrength::Bits256)?;

// Generate cryptographic random bytes
let mut key_material = [0u8; 32];
rng.fill_bytes(&mut key_material);

// Manual reseed for prediction resistance
rng.reseed()?;

Testing: 29 comprehensive tests validate FIPS compliance including:

• Known Answer Tests (KAT)
• Statistical distribution tests (chi-square)
• Continuous health monitoring
• Reseed mechanisms
• Non-repeatability validation

📚 API Documentation

Full API documentation is available at docs.rs/saorsa-pqc

Key Types

• MlKemPublicKey, MlKemSecretKey, MlKemCiphertext, MlKemSharedSecret
• MlDsaPublicKey, MlDsaSecretKey, MlDsaSignature
• SlhDsaPublicKey, SlhDsaSecretKey, SlhDsaSignature

Convenience Functions

• ml_kem_512(), ml_kem_768(), ml_kem_1024()
• ml_dsa_44(), ml_dsa_65(), ml_dsa_87()
• slh_dsa_sha2_128s(), slh_dsa_sha2_128f(), etc.

🛠️ Advanced Usage

Algorithm Selection Guide

Choose the right cryptographic primitives for your use case:

Hash Functions

use saorsa_pqc::api::hash::{Blake3Hasher, Sha3_256Hasher};
use saorsa_pqc::api::traits::Hash;

// High performance: BLAKE3
let mut hasher = Blake3Hasher::new();
hasher.update(b"data to hash");
let hash = hasher.finalize();

// NIST standard: SHA3-256
let mut hasher = Sha3_256Hasher::new();
hasher.update(b"data to hash");
let hash = hasher.finalize();

AEAD Encryption

use saorsa_pqc::api::aead::{Aes256GcmAead, AeadCipher, GcmNonce};
use saorsa_pqc::api::traits::Aead;

// Hardware accelerated: AES-256-GCM
let key = [0u8; 32]; // Use proper key generation
let aead = Aes256GcmAead::new(&key)?;
let nonce = GcmNonce::generate();
let ciphertext = aead.encrypt(&nonce, b"plaintext", b"aad")?;

// Software optimized: ChaCha20-Poly1305 (via enum)
let ciphertext = AeadCipher::ChaCha20Poly1305
    .encrypt(&key, nonce.as_ref(), b"plaintext", b"aad")?;

Key Derivation

use saorsa_pqc::api::kdf::HkdfSha3_256;
use saorsa_pqc::api::traits::Kdf;

// Derive encryption key from shared secret
let shared_secret = b"shared secret from ML-KEM";
let info = b"application context";
let mut derived_key = [0u8; 32];
HkdfSha3_256::derive(shared_secret, None, info, &mut derived_key)?;

HPKE (Hybrid Encryption)

use saorsa_pqc::api::hpke::{HpkeConfig, seal, open};
use saorsa_pqc::api::{MlKem, MlKemVariant, kdf::KdfAlgorithm, aead::AeadCipher};

// Configure HPKE with ML-KEM + AES-GCM
let config = HpkeConfig {
    kem: MlKemVariant::MlKem768,
    kdf: KdfAlgorithm::HkdfSha3_256,
    aead: AeadCipher::Aes256Gcm,
};

// Generate recipient keypair
let kem = MlKem::new(MlKemVariant::MlKem768);
let (pk, sk) = kem.generate_keypair()?;

// Encrypt
let (enc_key, ciphertext) = seal(
    config,
    &pk.to_bytes(),
    b"context info",
    b"secret message",
    b"associated data"
)?;

// Decrypt
let plaintext = open(
    config,
    &sk.to_bytes(),
    &enc_key,
    b"context info",
    &ciphertext,
    b"associated data"
)?;

Complete Quantum-Secure Communication

Combine ML-KEM key exchange with symmetric primitives:

use saorsa_pqc::api::{ml_kem_768, ChaCha20Poly1305};
use saorsa_pqc::api::symmetric::generate_nonce;

// Alice generates ML-KEM keypair
let kem = ml_kem_768();
let (alice_pk, alice_sk) = kem.generate_keypair()?;

// Bob encapsulates a shared secret using Alice's public key
let (shared_secret, ciphertext) = kem.encapsulate(&alice_pk)?;

// Alice decapsulates to get the same shared secret
let recovered_secret = kem.decapsulate(&alice_sk, &ciphertext)?;

// Derive proper encryption key from shared secret using HKDF
use saorsa_pqc::api::kdf::HkdfSha3_256;
use saorsa_pqc::api::traits::Kdf;

let mut encryption_key = [0u8; 32];
HkdfSha3_256::derive(
    &shared_secret.to_bytes(),
    None,
    b"saorsa-pqc encryption key",
    &mut encryption_key
)?;

// Create cipher with derived key
let cipher = ChaCha20Poly1305::new(&encryption_key);

// Now Bob can encrypt messages to Alice
let nonce = generate_nonce();
let message = b"Quantum-secure message";
let encrypted = cipher.encrypt(&nonce, message)?;

// Alice decrypts using the same key
let decrypted = cipher.decrypt(&nonce, &encrypted)?;
assert_eq!(decrypted, message);

🛠️ Additional Features

Serialization

// All keys and signatures support serialization
let pk_bytes = public_key.to_bytes();
let restored_pk = MlKemPublicKey::from_bytes(
    MlKemVariant::MlKem768, 
    &pk_bytes
)?;

Context Support (ML-DSA)

// ML-DSA supports domain separation via context
let context = b"application-specific-context";
let signature = dsa.sign_with_context(&secret_key, message, context)?;
let is_valid = dsa.verify_with_context(&public_key, message, &signature, context)?;

Deterministic Key Generation

// Generate keys from seed (for testing/reproducibility)
// Uses FIPS 203 deterministic generation with two 32-byte seeds
let d_seed = [0u8; 32];  // First seed value
let z_seed = [1u8; 32];  // Second seed value
let kem = ml_kem_768();
let (pk, sk) = kem.generate_keypair_from_seed(&d_seed, &z_seed);

// Deterministic generation produces identical keys
let (pk2, sk2) = kem.generate_keypair_from_seed(&d_seed, &z_seed);
assert_eq!(pk.to_bytes(), pk2.to_bytes());

🤝 Contributing

Contributions are welcome! Please feel free to submit issues and pull requests.

Development Setup

# Clone the repository
git clone https://github.com/dirvine/saorsa-pqc
cd saorsa-pqc

# Run tests
cargo test --all-features

# Run benchmarks
cargo bench

# Check code quality
cargo clippy --all-features
cargo fmt --check

📄 License

This project is dual-licensed under:

• MIT License
• Apache License 2.0

Choose whichever license works best for your use case.

🙏 Acknowledgments

This library builds upon the excellent work of:

• fips203 - ML-KEM implementation
• fips204 - ML-DSA implementation
• fips205 - SLH-DSA implementation
• blake3 - BLAKE3 hash function
• sha3 - SHA3 and Keccak implementations
• aes-gcm - AES-GCM AEAD cipher
• chacha20poly1305 - ChaCha20-Poly1305 AEAD
• hkdf - HMAC-based Key Derivation Function
• hmac - HMAC implementation

📮 Contact

• Author: David Irvine
• Email: [email protected]
• GitHub: @dirvine

🚀 Roadmap

• Hardware security module (HSM) support
• WebAssembly bindings
• C FFI bindings
• Hybrid modes (PQC + Classical)
• SHAKE256 XOF implementation
• Additional KDF algorithms (PBKDF2, Argon2)
• Side-channel resistance validation ✅ (v0.4.0 - DudeCT integration)
• Formal verification of critical paths
• Performance optimizations for specific platforms

---

📅 2024 NIST Updates

This library incorporates the latest NIST standards released in 2024:

• August 13, 2024: ML-KEM, ML-DSA, and SLH-DSA algorithms enabled on ACVTS Production server
• FIPS 203, 204, 205: Final standards published replacing draft versions
• Test Vectors: Updated to match the final NIST specifications

Note: This library is under active development. While the underlying FIPS implementations are certified, always perform your own security audit before production use.