Empower your software with autonomous capabilities. Convert legacy applications into agentic solutions or build new AI agents from scratch, all while leveraging a pre-built workflow engine that runs and abstracts your agent development process.
Install the AgentOrc SDK using your preferred package manager
1npm install @agentorc/sdk1pip install agentorc-sdkUse our SDK to wrap your existing application functionalities into autonomous agents. With just a few lines of code, turn any legacy system into a modular, self‑operating agent.
Leverage our Universal Data Envelope (UDE) to encapsulate all inputs and output, ensuring your agent can interoperate seamlessly within your application.
No need to rewrite your entire codebase. Our SDK acts as an abstraction layer that plugs directly into your existing system.
Our developer-friendly CLI tool and comprehensive SDK allow you to create new agents from scratch in minutes using pre-defined templates.
Test your agent locally using our simulation environment that mimics real-world orchestration. Validate your schemas and behavior before publishing.
Define your agent's capabilities, schemas, and parameters easily while adhering to standardized protocols for instant interoperability.
Publish your agent on the AgentOrc marketplace with a single API call. We handle all the metadata and performance metrics automatically.
Your agent becomes instantly visible to users seeking end-to-end automation, with rich performance and compatibility data.
End users install your agent with one click, with our platform managing authentication and configuration seamlessly.
Our SDK includes a robust workflow engine that abstracts away the complexities of asynchronous event processing and orchestration.
Register your agent functions as event handlers and let our engine manage retries, idempotency, and error handling automatically.
Our architecture automatically manages load balancing and auto-scaling, so you can focus on building business logic.
When you install the AgentOrc SDK—available for both Python and TypeScript—you receive a lightweight, modular toolkit that abstracts the complexities of creating agents and agent orchestration.
Every piece of data is wrapped in our Universal Data Envelope, ensuring consistent validation against strict schemas regardless of data type.
Define your agent's capabilities with our structured API, complete with pre-defined input and output schemas for automatic validation.
Register functions to handle specific events using our simple, declarative API. Our workflow engine manages routing and execution.
1import { AgentOrc, Agent, Event } from '@agentorc/sdk';
2
3// Initialize the SDK
4const agentorc = new AgentOrc({
5 apiKey: 'your_api_key'
6});
7
8// Define your agent
9@Agent({
10 name: 'data-processor',
11 description: 'Processes and analyzes data'
12})
13class DataProcessor {
14 @Event('analyze_data')
15 async analyzeData(data: any) {
16 // Your processing logic here
17 return {
18 result: 'processed data'
19 };
20 }
21}
22
23// Register your agent
24agentorc.register(DataProcessor);1from agentorc import AgentOrc, Agent, event
2
3# Initialize the SDK
4agentorc = AgentOrc(api_key='your_api_key')
5
6# Define your agent
7@Agent(
8 name='data-processor',
9 description='Processes and analyzes data'
10)
11class DataProcessor:
12 @event('analyze_data')
13 async def analyze_data(self, data):
14 # Your processing logic here
15 return {
16 'result': 'processed data'
17 }
18
19# Register your agent
20agentorc.register(DataProcessor)Get started with AgentOrc SDK in just a few simple steps
Download via npm (TypeScript) or pip (Python). Import and configure with your API key.
Use our templates and decorators to define your agent's capabilities and behavior.
Run your agent in our local simulation environment to validate schemas and behavior.