inuits_policy_based_auth
inuits_policy_based_auth is a python package for securing API endpoints based on policies.
Installation
Install inuits_policy_based_auth as follows:
pip install inuits-policy-based-auth
Getting Started
Instantiate the PolicyFactory in you app, for example in app.py (all examples given are based on a Python Flask app). The PolicyFactory requires a function that stores the user context for the duration of the request lifecycle. This responsibility is delegated to the API, so framework-specific implementations can handle more complex scenarios, such as multiprocess and multithreaded environments.
from flask import g
from inuits_policy_based_auth import PolicyFactory
def user_context_setter(user_context):
g.user_context = user_context
policy_factory = PolicyFactory(user_context_setter)
Manually loading policies
Importing and registering policies can be done manually. Don't forget to set a fallback key, this ensures that policies of a specific app are applied when the app cannot be determined automatically.
from inuits_policy_based_auth.authentication.policies.token_based_policies.authlib_flask_oauth2_policy import (
AuthlibFlaskOauth2Policy,
)
from inuits_policy_based_auth.authorization.policies.super_admin_policy import (
SuperAdminPolicy,
)
policy_factory.register_authentication_policy("apps.[app_name]", AuthlibFlaskOauth2Policy(...))
policy_factory.register_authorization_policy("apps.[app_name]", SuperAdminPolicy())
policy_factory.set_fallback_key_for_policy_mapping("apps.[app_name]")
However, it is recommended to load policies dynamically as this will allow you to make use of the full potential of this package.
Dynamically loading policies
You can write a loader which loads policies dynamically based on a configuration file.
Example configuration file:
{
"[app_name_1]": {
"policies": {
"authentication": [
"token_based_policies.authlib_flask_oauth2_policy",
"token_based_policies.tenant_token_roles_policy"
],
"authorization": [
"super_admin_policy",
"scope_based_policy"
]
}
},
"[app_name_2]": {
"policies": {
"authentication": [
"token_based_policies.authlib_flask_oauth2_policy",
"token_based_policies.tenant_token_roles_policy"
],
"authorization": [
"super_admin_policy",
"open_data_policy"
]
}
}
}
Example policy_loader.py:
import json
import os
from importlib import import_module
from inuits_policy_based_auth import PolicyFactory
from inuits_policy_based_auth.exceptions import (
PolicyFactoryException,
)
from logging import Logger
def load_policies(policy_factory: PolicyFactory, logger: Logger):
apps = {}
with open(str(os.getenv("[CONFIGURATION_FILE_NAME]")), "r") as configuration_file:
apps = json.load(configuration_file)
for app in apps:
try:
auth_type = "authentication"
for policy_module_name in apps[app]["policies"].get(auth_type):
policy = __get_class(app, auth_type, policy_module_name)
policy = __instantiate_authentication_policy(
policy_module_name, policy, logger
)
policy_factory.register_authentication_policy(
f"apps.{app}", policy
)
auth_type = "authorization"
for policy_module_name in apps[app]["policies"].get(auth_type):
policy = __get_class(app, auth_type, policy_module_name)
policy_factory.register_authorization_policy(
f"apps.{app}", policy()
)
except Exception as error:
raise PolicyFactoryException(
f"Policy factory was not configured correctly: {str(error)}"
).with_traceback(error.__traceback__)
policy_factory.set_fallback_key_for_policy_mapping("apps.[app_name]")
def __get_class(app, auth_type, policy_module_name):
locations = [
policy_module_name,
f"apps.{app}.policies.{auth_type}.{policy_module_name}",
f"inuits_policy_based_auth.{auth_type}.policies.{policy_module_name}",
]
for location in locations:
try:
module = import_module(location)
break
except ModuleNotFoundError:
pass
else:
raise ModuleNotFoundError(f"Policy {policy_module_name} not found")
policy_class_name = module.__name__.split(".")[-1].title().replace("_", "")
policy = getattr(module, policy_class_name)
return policy
def __instantiate_authentication_policy(policy_module_name, policy, logger: Logger):
token_schema = __load_token_schema()
if policy_module_name == "token_based_policies.authlib_flask_oauth2_policy":
allow_anonymous_users = (
True
if os.getenv("ALLOW_ANONYMOUS_USERS", "false").lower() == "true"
else False
)
return policy(
logger,
token_schema,
os.getenv("STATIC_ISSUER"),
os.getenv("STATIC_PUBLIC_KEY"),
os.getenv("ALLOWED_ISSUERS"),
allow_anonymous_users,
)
if policy_module_name == "token_based_policies.tenant_token_roles_policy":
return policy(
token_schema,
os.getenv("ROLE_SCOPE_MAPPING", os.getenv("API_SCOPES")),
True
if os.getenv("ALLOW_ANONYMOUS_USERS", "false").lower() == "true"
else False,
)
return policy()
def __load_token_schema() -> dict:
token_schema_path = os.getenv(
"TOKEN_SCHEMA", "path/to/token_schema.json"
)
with open(token_schema_path, "r") as token_schema:
return json.load(token_schema)
Now you can import the loader in app.py and pass policy_factory as an argument to it.
from apps.policy_loader import load_policies
from logging import Logger
load_policies(policy_factory, Logger(""))
As you can see in these examples, dynamically loading policies will allow you to add new policies and override existing ones, which makes this package highly customizable and generic.
Custom policies
Continuing from the examples above, you can make a custom authorization policy by creating a folder policies within a specific app. Here you should create the folders authentication and authorization that will contain custom policies which you can add to your configuration. In this case we name our new policy the same as an existing one, which will override it. Each authentication policy must inherit from BaseAuthenticationPolicy and implement the abstract method authenticate, while each authorization policy must inherit from BaseAuthorizationPolicy and implement the abstract method authorize.
Example folder structure:
api
├── apps
│ ├── [app_name]
│ │ ├── policies
│ │ │ ├── authentication
│ │ │ └── authorization
│ │ │ └── open_data_policy.py
│ │ ├── ...
...
│ ├── configuration.json
│ └── policy_loader.py
...
└── app.py
Example custom open_data_policy.py:
from inuits_policy_based_auth import BaseAuthorizationPolicy
class OpenDataPolicy(BaseAuthorizationPolicy):
def authorize(self, policy_context, _, request_context):
request = request_context.http_request
if request.method == "GET":
policy_context.access_verdict = True
return policy_context
Usage
If everything is set up correctly, you can use the apply_policies decorator as follows:
from app import policy_factory
from flask import request
from inuits_policy_based_auth import RequestContext
class Entity():
@policy_factory.apply_policies(
RequestContext(request, ["[scope_1]", "[scope_2]"])
)
def get(self):
...
You can also use the authenticate decorator to only apply authentication policies:
from app import policy_factory
from flask import request
from inuits_policy_based_auth import RequestContext
class Entity():
@policy_factory.authenticate(RequestContext(request))
def get(self):
...
Reusing an already authenticated user context
By default, apply_policies applies the authentication policies first and the authorization policies afterwards. If the user is already authenticated earlier in the same request, you can pass a user_context_getter to apply_policies. It is a function without arguments that returns the user context of the current request. When it returns a user context, the authentication policies are skipped and only the authorization policies are applied. When it returns None, the authentication policies are applied as usual.
Decorators are applied bottom-up, so put authenticate on the outside. It runs first and sets the user context, which apply_policies then reads through the getter.
from app import policy_factory
from flask import g, request
from inuits_policy_based_auth import RequestContext
def get_user_context():
return g.get("user_context")
class Entity():
@policy_factory.authenticate(RequestContext(request))
@policy_factory.apply_policies(
RequestContext(request, ["[scope_1]", "[scope_2]"]),
user_context_getter=get_user_context,
)
def patch(self):
...
The authenticate decorator accepts the same user_context_getter argument. This is useful when an endpoint decorated with authenticate is called from another endpoint that has already authenticated the user within the same request. If the getter returns a user context, the authentication policies are skipped and the endpoint is executed with that user context.
class Entity():
@policy_factory.authenticate(
RequestContext(request),
user_context_getter=get_user_context,
)
def patch(self):
...
Warning: the getter is called on every request, not when the endpoint is defined. Only use it if the user context comes from a trusted earlier step within the same request, such as the authenticate decorator. Never return a static or shared user context, as this would skip authentication for every request.
Contributing
Do not hesitate to open issues and create pull requests.
Metadata
Release files for inuits-policy-based-auth 10.3.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| inuits_policy_based_auth-10.3.0.tar.gz | 44.3 kB | With provenance | Details for inuits_policy_based_auth-10.3.0.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| inuits_policy_based_auth-10.3.0-py3-none-any.whl | Python 3 | none | any | Details for inuits_policy_based_auth-10.3.0-py3-none-any.whl |
Total release size: 99.4 kB
Release files / inuits_policy_based_auth-10.3.0.tar.gz
| Download URL | inuits_policy_based_auth-10.3.0.tar.gz |
|---|---|
| Size | 44.3 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
ce8142b74a5b08515e70b2186215e65d853923aff3e8cf1fb8e445fd0872b620
|
|
BLAKE2b-256 checksum How to use checksums |
7a1a098c6eab90ca1ce4eef7a19b9499d795e980a67007828ef024c34e39f166
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/7.0.0 CPython/3.13.16
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Nov 9, 2026.
Transparency logRelease files / inuits_policy_based_auth-10.3.0-py3-none-any.whl
| Download URL | inuits_policy_based_auth-10.3.0-py3-none-any.whl |
|---|---|
| Size | 55.1 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
039f5c62b3ac30362cfd430d31461336afa894097e57c0ca8b0919f593bc9c28
|
|
BLAKE2b-256 checksum How to use checksums |
1e5ce0a4f1506cedcd6d26f1458b7ffeccc29d0d9e306ff5b664f18ee4477223
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/7.0.0 CPython/3.13.16
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Nov 9, 2026.
Transparency logSecurity for inuits-policy-based-auth 10.3.0
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
Full provenance
Both files in this release have been uploaded with provenance metadata.
Attestation for inuits_policy_based_auth-10.3.0.tar.gz
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Nov 9, 2026.
Transparency logA note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.