Skip to main content

Parse and Process AWS IAM Policies, Statements, ARNs, and wildcards.

Project description

PolicyUniverse

Version

Build Status

Updater Status

Coverage Status

Code style: black

This package provides classes to parse AWS IAM and Resource Policies.

Additionally, this package can expand wildcards in AWS Policies using permissions obtained from the AWS Policy Generator.

See the Service and Permissions data.

This package can also minify an AWS policy to help you stay under policy size limits. Avoid doing this if possible, as it creates ugly policies. 💩

Install:

pip install policyuniverse

Usage:

Reading ARNs

from policyuniverse.arn import ARN
arn = ARN('arn:aws:iam::012345678910:role/SomeTestRoleForTesting')
assert arn.error == False
assert arn.tech == 'iam'
assert arn.region == ''  # IAM is universal/global
assert arn.account_number == '012345678910'
assert arn.name == 'role/SomeTestRoleForTesting'
assert arn.partition == 'aws'
assert arn.root == False  # Not the root ARN
assert arn.service == False  # Not an AWS service like lambda.amazonaws.com

arn = ARN('012345678910')
assert arn.account_number == '012345678910'

arn = ARN('lambda.amazonaws.com')
assert arn.service == True
assert arn.tech == 'lambda'

IAM and Resource Policies

Policy with multiple statements

# Two statements, both with conditions
policy05 = dict(
    Version='2010-08-14',
    Statement=[
        dict(
            Effect='Allow',
            Principal='arn:aws:iam::012345678910:root',
            Action=['s3:*'],
            Resource='*',
            Condition={
                'IpAddress': {
                    'AWS:SourceIP': ['0.0.0.0/0']
                }}),
        dict(
            Effect='Allow',
            Principal='arn:aws:iam::*:role/Hello',
            Action=['ec2:*'],
            Resource='*',
            Condition={
                'StringLike': {
                    'AWS:SourceOwner': '012345678910'
                }})
        ])

from policyuniverse.policy import Policy
from policyuniverse.statement import ConditionTuple, PrincipalTuple

policy = Policy(policy05)
assert policy.whos_allowed() == set([
    PrincipalTuple(category='principal', value='arn:aws:iam::*:role/Hello'),
    PrincipalTuple(category='principal', value='arn:aws:iam::012345678910:root'),
    ConditionTuple(category='cidr', value='0.0.0.0/0'),
    ConditionTuple(category='account', value='012345678910')
])

# The given policy is not internet accessible.
# The first statement is limited by the principal, and the condition is basically a no-op.
# The second statement has a wildcard principal, but uses the condition to lock it down.
assert policy.is_internet_accessible() == False

Internet Accessible Policy:

# An internet accessible policy:
policy01 = dict(
    Version='2012-10-08',
    Statement=dict(
        Effect='Allow',
        Principal='*',
        Action=['rds:*'],
        Resource='*',
        Condition={
            'IpAddress': {
                'AWS:SourceIP': ['0.0.0.0/0']
            }
        }))

policy = Policy(policy01)
assert policy.is_internet_accessible() == True
assert policy.internet_accessible_actions() == set(['rds:*'])

Statements

A policy is simply a collection of statements.

statement12 = dict(
    Effect='Allow',
    Principal='*',
    Action=['rds:*'],
    Resource='*',
    Condition={
        'StringEquals': {
            'AWS:SourceVPC': 'vpc-111111',
            'AWS:Sourcevpce': 'vpce-111111',
            'AWS:SourceOwner': '012345678910',
            'AWS:SourceAccount': '012345678910'
        },
        'StringLike': {
            'AWS:userid': 'AROAI1111111111111111:*'
        },
        'ARNLike': {
            'AWS:SourceArn': 'arn:aws:iam::012345678910:role/Admin'
        },
        'IpAddressIfExists': {
            'AWS:SourceIP': [
                '123.45.67.89',
                '10.0.7.0/24',
                '172.16.0.0/16']
        }
    })

from policyuniverse.statement import Statement
from policyuniverse.statement import ConditionTuple, PrincipalTuple

statement = Statement(statement12)
assert statement.effect == 'Allow'
assert statement.actions == set(['rds:*'])

# rds:* expands out to ~88 individual permissions
assert len(statement.actions_expanded) == 88

assert statement.uses_not_principal() == False
assert statement.principals == set(['*'])
assert statement.condition_arns == set(['arn:aws:iam::012345678910:role/Admin'])
assert statement.condition_accounts == set(['012345678910'])
assert statement.condition_userids == set(['AROAI1111111111111111:*'])
assert statement.condition_cidrs == set(['10.0.7.0/24', '172.16.0.0/16', '123.45.67.89'])
assert statement.condition_vpcs == set(['vpc-111111'])
assert statement.condition_vpces == set(['vpce-111111'])
assert statement.is_internet_accessible() == False
assert statement.whos_allowed() == set([
    PrincipalTuple(category='principal', value='*'),
    ConditionTuple(category='cidr', value='123.45.67.89'),
    ConditionTuple(category='account', value='012345678910'),
    ConditionTuple(category='userid', value='AROAI1111111111111111:*'),
    ConditionTuple(category='vpc', value='vpc-111111'),
    ConditionTuple(category='arn', value='arn:aws:iam::012345678910:role/Admin'),
    ConditionTuple(category='cidr', value='172.16.0.0/16'),
    ConditionTuple(category='vpce', value='vpce-111111'),
    ConditionTuple(category='cidr', value='10.0.7.0/24')])

Action Categories

policy = {
        "Statement": [{
            "Action": ["s3:put*", "sqs:get*", "sns:*"],
            "Resource": "*",
            "Effect": "Allow"
          }]
      }

from policyuniverse.policy import Policy
p = Policy(policy)
for k, v in p.action_summary().items():
    print(k,v)
>>> ('s3', set([u'Write', u'Permissions', u'Tagging']))
>>> ('sqs', set([u'List']))
>>> ('sns', set([u'List', u'Read', u'Write', u'Permissions']))

Possible categories are Permissions, Write, Read, Tagging, and List. This data can be used to summarize statements and policies and to look for sensitive permissions.

Expanding and Minification

from policyuniverse.expander_minimizer import expand_policy
from policyuniverse.expander_minimizer import minimize_policy

policy = {
        "Statement": [{
            "Action": ["swf:res*"],
            "Resource": "*",
            "Effect": "Allow"
          }]
      }

expanded_policy = expand_policy(policy=policy)
>>> Start size: 131. End size: 286
print(expanded_policy == {
        "Statement": [{
            "Action": [
              "swf:respondactivitytaskcanceled",
              "swf:respondactivitytaskcompleted",
              "swf:respondactivitytaskfailed",
              "swf:responddecisiontaskcompleted"
            ],
            "Resource": "*",
            "Effect": "Allow"
          }]
      })
>>> True

minimized_policy = minimize_policy(policy=expanded_policy, minchars=3)
>>> Skipping prefix r because length of 1
>>> Skipping prefix re because length of 2
>>> Skipping prefix r because length of 1
>>> Skipping prefix re because length of 2
>>> Skipping prefix r because length of 1
>>> Skipping prefix re because length of 2
>>> Skipping prefix r because length of 1
>>> Skipping prefix re because length of 2
>>> Start size: 286. End size: 131

print(minimized_policy == policy)
>>> True

Project details


Release history Release notifications | RSS feed

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

policyuniverse-1.5.0.20220523.tar.gz (387.1 kB view details)

Uploaded Source

Built Distribution

policyuniverse-1.5.0.20220523-py2.py3-none-any.whl (399.3 kB view details)

Uploaded Python 2 Python 3

File details

Details for the file policyuniverse-1.5.0.20220523.tar.gz.

File metadata

File hashes

Hashes for policyuniverse-1.5.0.20220523.tar.gz
Algorithm Hash digest
SHA256 826705f0a77018b314e60d4d620c4b2a004b935c89ad68bf7695444c3698d15a
MD5 e89466b4c782e1931613fa1a30a030b9
BLAKE2b-256 c7981cfb0d26a098ae695c6b004785008ff4d072dd75e1af5f37ed2ae485bd65

See more details on using hashes here.

File details

Details for the file policyuniverse-1.5.0.20220523-py2.py3-none-any.whl.

File metadata

File hashes

Hashes for policyuniverse-1.5.0.20220523-py2.py3-none-any.whl
Algorithm Hash digest
SHA256 997db60c3c0181a3fbae09e73a56f5c28076fa5e9b13ea09f93eedf9a0978fa7
MD5 51e89d729ef217200ac417c6c66a0945
BLAKE2b-256 31938f4f383d3553ad7f919a1001db4b67d0c49496ea29fec3f3871c093389f9

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page