AZTP Client Documentation

Enterprise-grade identity service client for secure workload identity management using AZTP standards.

(opens in a new tab) (opens in a new tab) (opens in a new tab) (opens in a new tab)

What is AZTP?

AZTP (Agentic Zero Trust Protocol) is a comprehensive security framework designed specifically for AI agent ecosystems and Model Context Protocol (MCP) implementations. It provides enterprise-grade identity management, policy enforcement, and secure communication protocols for distributed AI workloads.

Key Features

🔐 Zero Trust Identity Management - Cryptographic identity verification for AI components
📋 Policy-Based Access Control - Fine-grained permissions and organizational policies
🔗 Identity Flows & Linking - Manage complex multi-component workflows
🛡️ SAFE-MCP Protocol Support - Advanced security for MCP implementations
🌐 Multi-Language Support - Python and TypeScript/JavaScript clients
🏢 Enterprise Integration - OIDC, SPIFFE, and OAuth 2.0 compliance

Quick Navigation

Getting Started

Overview - Core concepts and architecture
Installation - Setup for Python and TypeScript
Quick Start - Your first secure connection
Trusted Domains - Understanding trust boundaries

Core Features

Identity Management - Create and manage workload identities
Policy Management - Access control and permissions
Real-World Examples - Production implementations

Advanced Topics

API Reference - Complete method documentation
Monitoring & Logging - Observability and debugging

Real-World Implementation

AZTP Client powers secure AI agent orchestration in production environments, including:

Langflow Integration - Secure component identity management in visual AI workflows
SAFE-MCP Protocol - Advanced security for Model Context Protocol implementations
Enterprise AI Platforms - Zero-trust architecture for distributed AI systems

Get Started

Tested and Verified Examples (September 10, 2025)

# Python - Real working example from test execution
import asyncio
import json
import os
from aztp_client import Aztp, whiteListTrustDomains
from dotenv import load_dotenv
 
load_dotenv()
 
async def get_started_example():
    """Real working AZTP example - tested September 10, 2025."""
    
    # Check available trusted domains
    print("🌐 Available Trusted Domains:")
    for domain_key, domain_value in whiteListTrustDomains.items():
        print(f"   • {domain_key}: {domain_value}")
    
    # Initialize client
    api_key = os.getenv("AZTP_API_KEY")
    client = Aztp(api_key=api_key)
    print("✅ AZTP client initialized successfully")
    
    # Create secure connection
    agent = await client.secure_connect({}, "demo-service", config={"isGlobalIdentity": False})
    
    # Get identity details
    identity_info = await client.get_identity(agent)
    identity_data = json.loads(identity_info)
    aztp_id = identity_data["data"]["aztpId"]
    
    print(f"🔑 Identity created: {aztp_id}")
    print(f"📊 Status: {identity_data['data']['status']}")
    print(f"🌍 Environment: {identity_data['data']['environment']}")
    print(f"🏠 Trust Domain: {identity_data['data']['workloadInfo']['trustDomain']}")
    
    # Example output:
    # 🔑 Identity created: aztp://astha.ai/workload/production/node/demo-service
    # 📊 Status: active
    # 🌍 Environment: production
    # 🏠 Trust Domain: astha.ai
    
    # Verify identity
    is_valid = await client.verify_identity(agent)
    print(f"🔍 Verification: {'✅ PASSED' if is_valid else '❌ FAILED'}")
    
    return agent
 
# Run: python get_started.py
if __name__ == "__main__":
    asyncio.run(get_started_example())

// TypeScript/JavaScript
import { AztpClient } from 'aztp-client';
 
const client = new AztpClient('your-api-key');
const agent = await client.secureConnect({}, 'demo-service', { isGlobalIdentity: false });
 
console.log(`Identity created: ${agent.identity.aztpId}`);

Test Execution Results Summary

✅ Successfully Tested on September 10, 2025:

19 Core Methods - All AZTP Client v1.0.42 methods tested and verified
2 Identities Created - Component identities with full lifecycle management
3 Identity Flows - Development, production, and testing workflows
Policy Evaluation - OIAP policies with real user scenarios
Flow-Specific Operations - Advanced flow management and revocation

Ready to secure your AI agents? Start with our Quick Start Guide.

Overview