Klira AI SDK Documentation
Welcome to the Klira AI SDK documentation. Klira AI is a powerful SDK for building AI applications with built-in guardrails, telemetry, and policy enforcement.
Getting Started
Overview
High-level introduction to Klira AI SDK
Installation
Setting up Klira AI SDK in your project
Quick Start
Build your first app with Klira AI SDK
First Example
Detailed walkthrough with explanations
Configuration
Configure Klira AI SDK for your needs
Core Concepts
Architecture
System architecture and design principles
Framework Detection
How automatic framework detection works
Adapters
Adapter system and framework integration
Observability
OpenTelemetry integration and tracing
Guardrails Overview
Introduction to policy enforcement system
Decorators Overview
Introduction to Klira AI decorators
Framework Integration
OpenAI Agents
OpenAI Agents SDK integration
LangChain
LangChain framework integration
CrewAI
CrewAI multi-agent framework integration
LlamaIndex
LlamaIndex query engine integration
LangGraph
LangGraph state machine integration
Multi-Framework
Using multiple frameworks together
Custom Frameworks
Integrating custom frameworks
Decorators
@workflow
Decorator for high-level processes
@task
Decorator for individual operations
@agent
Decorator for autonomous components
@tool
Decorator for utility functions
@crew
Decorator for CrewAI-specific workflows
@guardrails
Decorator for policy enforcement
@mcp_guardrails
Model Context Protocol protection
Decorator Composition
Combining multiple decorators
Policies
Custom policy decorators
LLM Adapters
OpenAI
OpenAI API integration
OpenAI Completions
Chat Completions API adapter
OpenAI Agents SDK
Agents SDK specific adapter
Anthropic
Anthropic Claude integration
Gemini
Google Gemini integration
Ollama
Local Ollama integration
Custom LLM
Custom LLM provider integration
Guardrails
Overview
Multi-layer guardrails architecture
Using Guardrails
Practical implementation guide
Policy Enforcement
Policy decision flow and enforcement levels
Configuration
Advanced guardrails configuration
MCP Protection
Model Context Protocol security policies
Monitoring
Performance tracking and real-time monitoring
Observability
Tracing
Distributed tracing with OpenTelemetry
Analytics
Event collection and analytics framework
Streaming Observability
Real-time streaming monitoring
Cache Analytics
Redis-backed distributed caching analytics
Advanced Configuration
Production Setup
Enterprise deployment configuration
Environment Variables
Complete environment variable reference
Performance Tuning
Cache optimization and performance settings
Security
API key management and content privacy
Scaling
Multi-instance deployment and auto-scaling
API Reference
API Overview
API reference navigation and overview
Klira Class
Core Klira class methods and initialization
Decorators API
Complete decorators API documentation
Guardrails API
GuardrailsEngine comprehensive API
Configuration API
Configuration management API
Types
Type definitions, protocols, and data structures
Utilities
Framework detection and utility functions