Overview

The GL AI Agents Platform (AIP) SDK is a comprehensive Python software development kit that provides developers with a simple, powerful interface to build and manage AI agents. It's designed to make AI agent development accessible, efficient, and enjoyable.

🎯 Value Proposition

  • Simple: Client() → create resources → agent.run("...")

  • Powerful: Full lifecycle management for agents, tools, and MCPs

  • Production-ready: Enterprise-grade with security and scalability

  • No telemetry: Your data stays private and local

👥 Who is this for?

  • Python developers building AI agents and applications

  • SREs and operations teams using the CLI in CI/CD pipelines

  • Platform teams integrating MCPs and custom tools

  • Data scientists creating AI-powered workflows

  • Startups and enterprises building production AI systems

🏗️ How it works

The SDK provides a clean abstraction layer that handles the complexity of AI agent management:

Core Components

  1. Python SDK: Clean, type-safe interface for programmatic access

  2. CLI Tool: Rich terminal interface for operations and scripting

  3. Resource Management: Unified handling of agents, tools, and MCPs

  4. Streaming Support: Real-time output with multiple renderer options

🎨 Design Principles

  • Simplicity First: Minimal boilerplate, maximum productivity

  • Consistency: Unified patterns across all resource types

  • Safety: Type validation, error handling, and secure defaults

  • Observability: Rich logging, debugging, and monitoring support

🌍 Supported Environments

  • Operating Systems: Windows, macOS, Linux

  • Python Versions: 3.10+

  • Deployment: Local development, cloud platforms, containers

  • Integrations: LangChain, custom frameworks, external APIs

🚀 Key Capabilities

Agent Management

  • Create, configure, and run AI agents

  • Support for multiple language models

  • Agent-to-agent communication

  • Memory and context management

Tool Integration

  • Custom Python tool creation

  • Framework support (LangChain, custom)

  • File upload and processing

  • External service integration

Note: All tools must contain actual Python code with the @tool_plugin decorator and inherit from BaseTool. The CLI supports file uploads via --file option.

MCP Support

  • Model Context Protocol integration

  • Universal API adapter

  • Service discovery and testing

  • Tool enumeration and management

CLI Operations

  • Rich terminal interface

  • Shell completion support

  • Scripting-friendly output formats

  • Configuration management

💡 Developer Benefits

  • Fast Onboarding: Get your first agent running in under 5 minutes

  • Productive Development: Intuitive APIs that feel natural

  • Robust Error Handling: Clear error messages and recovery paths

  • Comprehensive Testing: Built-in validation and testing support

  • Production Ready: Enterprise-grade security and scalability

🔒 Security & Privacy

  • No Telemetry: Zero data collection or tracking

  • Secure Communication: TLS encryption for all API calls

  • Credential Management: Secure handling of API keys and tokens

  • Local Processing: Your data stays on your infrastructure

📈 Performance & Scalability

  • Efficient HTTP: Connection pooling and session reuse

  • Resource Management: Automatic cleanup and lifecycle management

  • Timeout Handling: Configurable timeouts for long-running operations

  • Error Recovery: Robust retry mechanisms and error handling

🎯 Use Cases

  • AI Chatbots: Intelligent conversational agents

  • Task Automation: Complex workflow automation

  • Data Analysis: AI-powered data processing

  • API Integration: External service orchestration

  • Custom Tools: Domain-specific AI capabilities

  • Multi-Agent Systems: Coordinated AI networks

🚀 Next Steps

Ready to get started? Jump to Quick Start to create your first AI agent in under 5 minutes!

The AIP SDK transforms complex AI agent development into simple, enjoyable programming. Start building today!

PreviousA2ANextGet Started