autoagentsai 0.1.36

SDK for interacting with Autoagents.ai API

Professional AI Platform Python SDK

English | 简体中文

Professional Python SDK for AutoAgents AI platform, providing intuitive APIs for intelligent conversation, file processing, knowledge base management, and more.

Table of Contents

• Why AutoAgents AI Python SDK?
• Quick Start
• Core Features
• API Reference
• Configuration
• Examples
• Contributing
• License

Why AutoAgents AI Python SDK?

AutoAgents AI Python SDK is a comprehensive toolkit that transforms how developers interact with AI-powered automation systems. Built for modern Python applications, it provides seamless integration with the AutoAgents AI platform.

Core Features

Intelligent Conversation

• Streaming Chat: Real-time conversation with multi-turn interactions
• Reasoning Process: Display AI thinking and decision-making steps
• Multi-modal Support: Handle text, images, and files in unified interface

File Processing

• Multi-format Support: Automatic processing of PDF, Word, images, and more
• Smart Analysis: Extract insights and content from documents
• Batch Operations: Handle multiple files efficiently

Knowledge Base Management

• Complete CRUD Operations: Create, read, update, delete knowledge bases
• Advanced Search: Semantic search and content retrieval
• Content Organization: Structured storage and management

Pre-built Agents

• PowerPoint Generation: Create presentations from templates and data
• React Agents: Interactive problem-solving agents
• Workflow Automation: Complex multi-step task orchestration
• Data Science Tools: Analytics and visualization capabilities

Modern Architecture

• Async Support: High-performance asynchronous API calls
• Type Safety: Full Pydantic type validation
• Extensible Design: Modular components for custom solutions

Why Choose AutoAgents AI Python SDK?

• Developer-First: Intuitive APIs designed for modern Python development
• Production-Ready: Battle-tested in enterprise environments
• Comprehensive: Everything needed for AI automation in one package
• Well-Documented: Extensive examples and clear API documentation

Quick Start

Prerequisites

• Python 3.11+
• AutoAgents AI platform account

Installation

pip install autoagentsai

Or install from source:

git clone https://github.com/your-repo/autoagents-python-sdk.git
cd autoagents-python-sdk
pip install -e .

Get API Keys

1. Log in to AutoAgents AI platform
2. Navigate to Profile → Personal Keys
3. Copy your personal_auth_key and personal_auth_secret

First Conversation

from autoagentsai.client import ChatClient

# Initialize client
client = ChatClient(
    agent_id="your_agent_id",
    personal_auth_key="your_auth_key", 
    personal_auth_secret="your_auth_secret"
)

# Start conversation
for event in client.invoke("Hello, please introduce artificial intelligence"):
    if event['type'] == 'token':
        print(event['content'], end='', flush=True)
    elif event['type'] == 'finish':
        break

File Processing

# Upload and analyze files
for event in client.invoke(
    prompt="Please analyze the main content of this document",
    files=["document.pdf"]
):
    if event['type'] == 'token':
        print(event['content'], end='', flush=True)

Knowledge Base Management

from autoagentsai.client import KbClient

# Initialize knowledge base client
kb_client = KbClient(
    personal_auth_key="your_auth_key",
    personal_auth_secret="your_auth_secret"
)

# Create knowledge base
result = kb_client.create_kb(
    name="Technical Documentation",
    description="Store technical documents"
)

# Query knowledge base list
kb_list = kb_client.query_kb_list()

Slide Generation

from autoagentsai.slide import SlideAgent

# Create slide agent
slide_agent = SlideAgent()

# Generate presentation
slide_agent.fill(
    prompt="Create a presentation about AI development",
    template_file_path="template.pptx",
    output_file_path="output.pptx"
)

Advanced Workflow Automation

from autoagentsai.graph import FlowGraph

# Create workflow graph
graph = FlowGraph(
    personal_auth_key="your_auth_key",
    personal_auth_secret="your_auth_secret"
)

# Add workflow nodes and compile
graph.add_node("chat_node", "chat", {"prompt": "Analyze this data"})
graph.add_node("ppt_node", "slide", {"template": "report.pptx"})
graph.add_edge("chat_node", "ppt_node")

# Deploy workflow
graph.compile(workflow_name="data_analysis_pipeline")

API Reference

ChatClient

Main conversation client supporting streaming chat and multimodal input.

Methods

• invoke(prompt, images=None, files=None) - Start conversation
• history() - Get conversation history

Event Types

• start_bubble - New response bubble starts
• token - Text fragment (for typewriter effect)
• reasoning_token - AI reasoning process
• end_bubble - Response bubble ends
• finish - Conversation complete

KbClient

Knowledge base management client.

Methods

• create_kb(name, description) - Create knowledge base
• query_kb_list() - Query knowledge base list
• get_kb_detail(kb_id) - Get knowledge base details
• delete_kb(kb_id) - Delete knowledge base

FlowGraph

Workflow automation and orchestration.

Methods

• add_node(node_id, module_type, inputs) - Add workflow node
• add_edge(source, target) - Connect nodes
• compile(workflow_name) - Deploy workflow

Configuration

Environment Settings

# Development environment (default)
base_url = "https://uat.agentspro.cn"

# Production environment
base_url = "https://agentspro.cn"

API Keys Setup

Set your credentials as environment variables:

export AUTOAGENTS_AUTH_KEY="your_auth_key"
export AUTOAGENTS_AUTH_SECRET="your_auth_secret"

Or pass them directly when initializing clients:

client = ChatClient(
    personal_auth_key="your_auth_key",
    personal_auth_secret="your_auth_secret"
)

Getting Agent ID

1. Open Agent details page
2. Click "Share" → "API"
3. Copy Agent ID

Examples

Explore the playground/ directory for comprehensive examples:

• playground/client/ - Chat and API examples
• playground/slide/ - PowerPoint generation examples
• playground/kb/ - Knowledge base management
• playground/react/ - React Agent examples
• playground/graph/ - Workflow automation
• playground/datascience/ - Data analysis tools

Contributing

We welcome contributions! Please feel free to submit issues and pull requests.

Development Setup

git clone https://github.com/your-repo/autoagents-python-sdk.git
cd autoagents-python-sdk
pip install -e .[dev]

License

MIT License

Support

• Email: [e-mail]
• Documentation: AutoAgents AI Official Docs
• Issues: GitHub Issues