policymesh 0.1.0

Python SDK for the PolicyMesh AI Agent Control Platform

PolicyMesh Python SDK

The official Python SDK for the PolicyMesh AI Agent Control Platform.

Agents propose. PolicyMesh enforces.

Installation

pip install policymesh

Quick Start

from policymesh import PolicyMeshClient

# Initialize the client with your org ID
client = PolicyMeshClient(org_id="your_org_id")

# Evaluate an agent action
decision = client.evaluate(
    agent_id="my_agent",
    action_type="data_export",
    data_classification="confidential",
    record_count=1500,
    description="Exporting customer records to external email"
)

print(decision.decision)      # "block"
print(decision.policy_matched) # "Sensitive Data Export Block"
print(decision.is_blocked)    # True

Using the Guard Decorator

from policymesh import PolicyMeshClient

client = PolicyMeshClient(org_id="your_org_id")

@client.guard(action_type="production_deploy", environment="production")
def deploy_to_production():
    # This function will be blocked if PolicyMesh says so
    print("Deploying to production...")

# If blocked, raises PolicyBlockedError before the function runs
deploy_to_production()

Simple Boolean Check

# Returns True if allowed, False if blocked
if client.allow(agent_id="my_agent", action_type="data_access"):
    # proceed with action
    pass

Handling Decisions

from policymesh import PolicyMeshClient, PolicyBlockedError, PolicyEscalateError

client = PolicyMeshClient(org_id="your_org_id")

try:
    decision = client.evaluate(
        agent_id="my_agent",
        action_type="production_deploy",
        environment="production"
    )
    if decision.is_flagged:
        print("Action flagged but allowed. Proceeding with caution.")
except PolicyBlockedError as e:
    print(f"Action blocked: {e.policy_matched}")
except PolicyEscalateError as e:
    print(f"Action requires approval: {e.policy_matched}")

Action Types

Action Type	Description
data_export	Exporting data to external systems
external_email	Sending emails to external addresses
production_deploy	Deploying to production environments
permission_change	Modifying user permissions or access
payment	Processing payments or vendor actions
data_access	Accessing sensitive data
custom	Custom action types

Data Classifications

Classification	Description
public	Publicly available data
internal	Internal company data
confidential	Confidential business data
restricted	Highly restricted or regulated data

Links

• Dashboard
• Documentation
• GitHub