pqjwt 0.2.1

# PQC JWT Python Library

PQC JWT Python Library

A comprehensive Python library for generating, managing, signing, and verifying Post-Quantum Cryptography (PQC) JSON Web Tokens (JWTs). Supports ML-DSA (Dilithium), Falcon, and SPHINCS+ digital signature algorithms via the pqcrypto package.

This library encapsulates key management and JWT logic, making it easy to integrate quantum-resistant signatures into your application's authentication flow.

---

Features

• Post-Quantum Ready: Implements NIST-standardized ML-DSA, Falcon padded, and SPHINCS+ signature algorithms for quantum-safe JWTs
• Multiple Algorithm Support: Handles multiple security levels and schemes:
  • ML-DSA (Dilithium): ML-DSA-44, ML-DSA-65, ML-DSA-87
  • Falcon Padded: Falcon-512, Falcon-1024
  • SPHINCS+: Multiple variants with SHA2 and SHAKE hashing
• Flexible Key Storage: Save/load key pairs in multiple formats:
  • pem: Standard-compliant ASN.1 structures (PKCS#8 for private keys, X.509 SPKI for public keys) with official NIST Object Identifiers (OIDs).
  • pub: Simple public key in hex format
  • bin: Raw binary key bytes
• Publisher/Consumer Roles:
  • Publisher: Generates keys and signs JWTs
  • Consumer: Loads public keys and verifies JWTs
• Standard JWT Claims Validation: Automatically validates exp, nbf, and iat claims
• Comprehensive Error Handling: Detailed exceptions for all error scenarios

---

Requirements

• Python 3.7 or higher
• pqcrypto package

pip install pqcrypto pyasn1

---

Installation

From PyPI

pip install pqjwt

From Source

git clone https://github.com/eduardogiraudi/pqjwt.git
cd pqjwt
pip install -e .

---

Quick Start

Basic Usage

import time
from pqjwt import create_publisher, create_consumer

# Publisher: generates keys if missing, signs JWTs
publisher = create_publisher(
    key_dir="./keys", 
    algorithm="ML-DSA-44", 
    key_format="pem"
)

# Create JWT payload with standard claims
payload = {
    "user_id": 123, 
    "role": "admin", 
    "iat": int(time.time()),  # Issued at
    "exp": int(time.time()) + 3600  # Expires in 1 hour
}

# Sign and encode JWT
jwt_token = publisher.encode(payload)
print("Generated JWT:", jwt_token)

# Consumer: loads public key, verifies JWT
consumer = create_consumer(
    key_dir="./keys", 
    algorithm="ML-DSA-44", 
    key_format="pem"
)

try:
    headers, claims = consumer.decode(jwt_token)
    print("JWT is valid!")
    print("Headers:", headers)
    print("Claims:", claims)
except Exception as e:
    print("Verification failed:", e)

Advanced Usage with Custom Headers

from pqjwt import create_publisher

publisher = create_publisher(algorithm="FN-DSA-512")

# Custom JWT headers
custom_headers = {
    "alg": "FN-DSA-512",
    "typ": "JWT", 
    "kid": "key-001"  # Key ID for key rotation
}

payload = {
    "sub": "[e-mail]",
    "scope": "read write admin",
    "exp": int(time.time()) + 7200
}

jwt_token = publisher.encode(payload, headers=custom_headers)
print("JWT with custom headers:", jwt_token)

---

Supported Algorithms

ML-DSA (Dilithium) - NIST Standardized

Algorithm	JWT Header	Security Level	Description
ML-DSA-44	ML-DSA-44	Level 2	Balanced security/performance
ML-DSA-65	ML-DSA-65	Level 3	Higher security
ML-DSA-87	ML-DSA-87	Level 5	Highest security

Falcon Padded - Timing Attack Resistant

Algorithm	JWT Header	Security Level	Signature Size
FN-DSA-512	FN-DSA-512	Level 1	~690 bytes
FN-DSA-1024	FN-DSA-1024	Level 5	~1330 bytes

SPHINCS+ - Hash-Based Security

Algorithm	JWT Header	Security Level	Signature Size
SLH-DSA-SHA2-128f	SLH-DSA-SHA2-128f	Level 1	~17KB
SLH-DSA-SHA2-128s	SLH-DSA-SHA2-128s	Level 1	~8KB
SLH-DSA-SHA2-192f	SLH-DSA-SHA2-192f	Level 3	~35KB
SLH-DSA-SHA2-192s	SLH-DSA-SHA2-192s	Level 3	~16KB
SLH-DSA-SHA2-256f	SLH-DSA-SHA2-256f	Level 5	~49KB
SLH-DSA-SHA2-256s	SLH-DSA-SHA2-256s	Level 5	~22KB

Plus SHAKE variants for all SPHINCS+ algorithms

---

Key Management

Key Formats

PEM Format (Default)

# Keys stored as:
# - ml-dsa-44_public.pem
# - ml-dsa-44_private.pem

publisher = create_publisher(
    key_dir="./keys",
    algorithm="ML-DSA-44", 
    key_format="pem"  # Default
)

Example PEM file:

-----BEGIN PUBLIC KEY-----
MOCK_BASE64_DATA_HERE
-----END PUBLIC KEY-----

PUB Format (Public Key Only)

# Public key stored as:
# - ml-dsa-44_public.pub

publisher = create_publisher(
    key_dir="./keys",
    algorithm="ML-DSA-44",
    key_format="pub"
)

Example PUB file:

ML-DSA-44 PUBLIC KEY
a1b2c3d4e5f6... (hex encoded key)

BIN Format (Raw Binary)

# Keys stored as raw binary:
# - ml-dsa-44_public.bin
# - ml-dsa-44_private.bin

publisher = create_publisher(
    key_dir="./keys",
    algorithm="ML-DSA-44",
    key_format="bin"
)

Manual Key Management

from pqjwt import JWTKeyManager

# Save keys manually
public_key = b"..."  # Raw public key bytes
JWTKeyManager.save_key(
    public_key, 
    "my_key.pem", 
    format_type="pem", 
    key_type="public",
    algorithm="ML-DSA-44"
)

# Load keys manually
loaded_key, algorithm = JWTKeyManager.load_key(
    "my_key.pem", 
    format_type="pem", 
    key_type="public"
)

---

API Reference

JWTManager Class

Main class for JWT operations.

Initialization

JWTManager(
    mode="publisher",      # "publisher" or "consumer"
    key_dir="./keys",      # Directory for key storage
    key_format="pem",      # "pem", "pub", or "bin"
    algorithm="ML-DSA-44"  # Any supported algorithm
)

Methods

encode(payload, headers=None)

Signs a payload and returns JWT string.

publisher = create_publisher()
jwt = publisher.encode(
    payload={"user": "123", "exp": ...},
    headers={"alg": "ML-DSA-44", "typ": "JWT"}
)

decode(jwt, validate_claims=True, clock_skew=5)

Verifies JWT and returns (headers, payload).

consumer = create_consumer()
headers, payload = consumer.decode(
    jwt_token,
    validate_claims=True,  # Validate exp, nbf, iat
    clock_skew=5           # Allow 5 seconds clock difference
)

verify(jwt)

Quick verification without full decoding.

is_valid = consumer.verify(jwt_token)

get_public_key(output_format="bytes")

Get public key in different formats.

key_bytes = publisher.get_public_key("bytes")
key_hex = publisher.get_public_key("hex") 
key_pem = publisher.get_public_key("pem")

export_public_key(file_path, format_type=None)

Export public key to file.

publisher.export_public_key("backup_key.pem", "pem")

Factory Functions

create_publisher(key_dir="./keys", key_format="pem", algorithm="ML-DSA-44")

Creates a JWTManager instance in publisher mode.

create_consumer(key_dir="./keys", key_format="pem", algorithm="ML-DSA-44")

Creates a JWTManager instance in consumer mode.

JWTKeyManager Class

Static utility methods for key management.

get_supported_algorithms()

Returns list of all supported algorithms.

algorithms = JWTKeyManager.get_supported_algorithms()
print("Supported:", algorithms)

get_jwt_header_name(algorithm)

Maps algorithm name to JWT header.

header = JWTKeyManager.get_jwt_header_name("ML-DSA-44")
print(header)  # "ML-DSA-44"

get_algorithm_from_jwt_header(jwt_header)

Maps JWT header back to algorithm name.

algorithm = JWTKeyManager.get_algorithm_from_jwt_header("ML-DSA-44")
print(algorithm)  # "ML-DSA-44"

---

Error Handling

The library provides detailed exceptions for all error scenarios:

from pqjwt import (
    JWTExpiredError,
    JWTSignatureError, 
    JWTValidationError,
    AlgorithmNotSupportedError,
    CryptoKeyError
)

try:
    headers, payload = consumer.decode(jwt_token)
except JWTExpiredError as e:
    print("Token expired:", e)
except JWTSignatureError as e:
    print("Invalid signature:", e)
except JWTValidationError as e:
    print("Validation failed:", e)
except AlgorithmNotSupportedError as e:
    print("Algorithm not supported:", e)

Available Exceptions

• JWTExpiredError: Token has expired
• JWTSignatureError: Signature verification failed
• JWTValidationError: General validation failure (malformed token, invalid claims)
• JWTDecodeError: Error decoding Base64 or JSON
• CryptoKeyError: Key-related errors (missing private key, etc.)
• AlgorithmNotSupportedError: Unsupported algorithm specified
• KeyFormatError: Invalid key format

---

Security Considerations

Algorithm Security

• ML-DSA: NIST Standardized (FIPS 204), no known practical attacks
• Falcon Padded: Mitigates timing attacks present in basic Falcon variants
• SPHINCS+: Conservative hash-based security, very large signatures but resistant to lattice attacks

Key Protection

• Store private keys securely with appropriate file permissions
• Never expose private keys in client applications
• Use different key directories for different environments

Best Practices

# Use appropriate security levels
production_algorithm = "ML-DSA-65"  # Level 3 security
testing_algorithm = "ML-DSA-44"     # Level 2 security

# Set reasonable expiration times
payload = {
    "exp": int(time.time()) + 3600,  # 1 hour for access tokens
    "iat": int(time.time())
}

---

Testing

Run the test suite:

python -m pytest tests/ -v

Or run the test file directly:

python tests/test.py

---

File Naming Convention

Keys are automatically named based on algorithm and format:

Algorithm	Format	Public Key File	Private Key File
ML-DSA-44	pem	ml-dsa-44_public.pem	ml-dsa-44_private.pem
FN-DSA-512	pub	fn-dsa-512.pub	fn-dsa-512_private.pem
SLH-DSA-SHA2-128f	bin	slh-dsa-sha2-128f_public.bin	slh-dsa-sha2-128f_private.bin

Note: Private keys are always stored in PEM format when using PUB format for public keys.

---

Contributing

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Pull Request

---

License

This project is licensed under the MIT License - see the LICENSE file for details.

---

Troubleshooting

Common Issues

"Algorithm not supported" Error

• Check algorithm name spelling
• Use JWTKeyManager.get_supported_algorithms() to see available options

"Public key not found" Error

• Ensure key files exist in the specified directory
• Check file naming convention matches algorithm

Signature Verification Fails

• Ensure same algorithm used for signing and verification
• Check clock skew for time-based claims

Performance Issues with SPHINCS+

• SPHINCS+ has large signature sizes, consider using ML-DSA or Falcon for high-throughput applications

---

Support

For bugs and feature requests, please open an issue on GitHub.

For security vulnerabilities, please contact the maintainers directly.