superjwt 0.4.0

A super simple JWT library for Python

A modern implementation of JSON Web Token (JWT) for Python.
With powerful Pydantic validation features.

See documentation

Overview & Installation

SuperJWT is a minimalist JWT library for Python 3.10+ that combines the simplicity of JWT encoding/decoding with the power of Pydantic validation. It supports JWS (JSON Web Signature) format with HMAC-SHA2 algorithms and includes advanced features like token inspection and detached payload mode.

Key Features:

• 🔐 Secure by default - JWS signature algorithm required.
• 🪶 Minimalist - Clean, modern code with minimal dependencies.
• ✔️ JWT validation - Easy claims validation with Pydantic models.
• 🏷️ Type hints - IDE autocompletion with your JWT claims or JOSE headers.

Install via pip:

pip install superjwt

---

Usage

SuperJWT makes it easy to encode and decode JWT tokens with automatic validation and serialization. Here are the fundamental operations:

Basic Usage 🐣

Encode manually your claims from a dict. During decoding, validate your JWT content against the standard JWT claims.

from superjwt import JWTClaims, encode, decode

secret_key = "your-secret-key-of-len-32-bytes!"

compact: bytes = encode({"iss": "my-app", "sub": "John Doe"}, secret_key, "HS256")
#> b'eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJpc3MiOiJteS1hcHAiLCJzdWIiOiJKb2huIERvZSJ9.HwnUqTLFAMzNkMrokd0aI7c-zSJJpSVXMrYIhUyWe4s'
decoded: dict = decode(compact, secret_key, "HS256", claims_validation=JWTClaims)
#> {'iss': 'my-app', 'sub': 'John Doe'}

Define dynamically your claims with Pydantic and easily include 'iat' (Issued At) and 'exp' (Expiration). Validate your JWT content automatically during encoding and decoding.

from superjwt import JWTClaims, encode, decode

secret_key = "your-secret-key-of-len-32-bytes!"

claims = (
    JWTClaims(iss="my-app", sub="John Doe")
    .with_issued_at()
    .with_expiration(minutes=15)
)

compact: bytes = encode(claims, secret_key, "HS256")
#> b'eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJpc3MiOiJteS1hcHAiLCJzdWIiOiJKb2huIERvZSIsImlhdCI6MTc2NzAyNzQ4MywiZXhwIjoxNzY3MDI4MzgzfQ.ZXxZT8VzL8IPTov-enslCh57S2M5fQBtqULZx5zEAm8'
decoded: dict = decode(compact, secret_key, "HS256", claims_validation=JWTClaims)
#> {'iss': 'my-app', 'sub': 'John Doe', 'iat': 1767027483, 'exp': 1767028383}

Custom Claims and Validation

Redefine standard claims or define new custom ones. Validate automatically during encoding and decoding.

from typing import Annotated
from uuid import UUID

from pydantic import AfterValidator, Field
from superjwt import JWTClaims, Validation, decode, encode
from superjwt.exceptions import ClaimsValidationError

secret_key = "your-secret-key-of-len-32-bytes!"

class MyJWTClaims(JWTClaims):
    # redefine 'sub' as required integer
    sub: int = Field(default=...)

    # new custom claim:  'user_id' is required and must be a valid UUIDv4 string
    user_id: Annotated[str, AfterValidator(lambda x: str(UUID(x, version=4)))]

# Example - Validation PASSING

claims = (
    MyJWTClaims(sub=123, user_id="b2a4c791-2cf4-4e41-9a20-8532129ff47c")
    .with_issued_at()
    .with_expiration(minutes=15)
)
compact = encode(claims, secret_key, "HS256")
decoded = decode(compact, secret_key, "HS256", claims_validation=MyJWTClaims)
#> {'sub': 123, 'iat': 1767026691, 'exp': 1767027591, 'user_id': 'b2a4c791-2cf4-4e41-9a20-8532129ff47c'}

# Example - Validation FAILING

# create a non-validated and invalid pydantic claims
invalid_claims = (
    MyJWTClaims.model_construct(**{"sub": "John Doe", "user_id": "invalid-uuid-string"})
    .with_issued_at()
    .with_expiration(minutes=10)
)

# disable claims default validation to create an invalid token
invalid_compact = encode(
    invalid_claims, secret_key, "HS256", claims_validation=Validation.DISABLE
)
try:
    decode(invalid_compact, secret_key, "HS256", claims_validation=MyJWTClaims)
except ClaimsValidationError as e:
    print("Claims validation error:", e)
#> Claims validation error: Claims validation failed
#> claim ('sub',) = John Doe -> validation failed (int_parsing): Input should be a valid integer, unable to parse string as an integer
#> claim ('user_id',) = invalid-uuid-string -> validation failed (value_error): Value error, badly formed hexadecimal UUID string

Documentation

See full documentation

Test

1. Clone repository

2. Install dependencies

   pip install -e . --group test
   

3. Run tests

   pytest