Sekuire SDKs

Build AI agents with the Sekuire Secure Layer using our official SDKs for TypeScript, Python, and Rust.

tip

All SDKs use the same sekuire.yml configuration file format. Use the CLI to scaffold projects.

Installation

TypeScript

npm install @sekuire/sdk
# or
pnpm add @sekuire/sdk

Python

pip install sekuire-sdk

Rust

# Add to Cargo.toml
[dependencies]
sekuire-sdk = "0.1"
tokio = { version = "1", features = ["full"] }

Quick Start

Create an agent with the CLI, then use the SDK to interact with it:

TypeScript

import { getAgent } from '@sekuire/sdk';

// Load agent from sekuire.yml
const agent = await getAgent('assistant');

// Chat with the agent
const response = await agent.chat('Hello!');
console.log(response);

Python

import asyncio
from sekuire_sdk import get_agent

async def main():
    # Load agent from sekuire.yml
    agent = await get_agent('assistant')

    # Chat with the agent
    response = await agent.chat('Hello!')
    print(response)

asyncio.run(main())

Rust

use sekuire_sdk::get_agent;

#[tokio::main]
async fn main() -> anyhow::Result<()> {
    // Load agent from sekuire.yml
    let mut agent = get_agent(Some("assistant"), None).await?;

    // Chat with the agent
    let response = agent.chat("Hello!", None).await?;
    println!("{}", response);

    Ok(())
}

Features

Feature	TypeScript	Python	Rust
LLM Providers	OpenAI, Anthropic, Google, Ollama	OpenAI, Anthropic, Google, Ollama	OpenAI, Anthropic, Google, Ollama
Config-First	Yes	Yes	Yes
Streaming	Yes	Yes	Yes
Built-in Tools	Yes	Yes	Yes
Memory	Buffer, Window	Buffer, Window	Buffer, Window
Type Safety	TypeScript	Type Hints	Rust types

Core Concepts

Config-First Approach

All SDKs read from sekuire.yml - the same config file used by the CLI. This ensures consistency between your development workflow and runtime.

LLM Providers

Switch providers by changing the config. All SDKs support:

• OpenAI - GPT-4, GPT-4 Turbo, GPT-3.5 Turbo
• Anthropic - Claude 3 Opus, Sonnet, Haiku
• Google - Gemini Pro, Gemini 1.5
• Ollama - Local models (Llama, Mistral, etc.)

Async by Default

All SDKs are async-first for optimal performance with I/O-bound operations.

API Reference

Core API

Command	Description
getAgent()	Load single agent from config
getAgents()	Load all agents from config
agent.chat()	Send message and get response
agent.chatStream()	Stream response tokens
Built-in Tools	Calculator, HTTP, Web Search

Platform Features

Command	Description
A2AClient / A2AServer	Agent-to-agent communication
createBeacon()	Heartbeat and kill switch
SekuireServer	HTTP server with trust endpoints
PolicyEnforcer	Runtime policy enforcement
createMemoryStorage()	Pluggable memory backends
initTelemetry()	OpenTelemetry integration
createWorker()	Background task processing
LLM Providers	OpenAI, Anthropic, Google, Ollama

Advanced Features

Feature	Description
Multi-Agent Orchestration	Discover and delegate across agents
Agent Identity	Cryptographic signing and verification
Logging	Compliance event logging

Reference

Topic	Description
Tools Reference	All 50+ built-in tools
Environment Variables	Configuration reference
Authentication	Identity and trust
Error Handling	Error types and recovery

Next Steps

• Quickstart Guide - Full walkthrough
• Agent API - Creating and using agents
• Built-in Tools - Calculator, HTTP, File I/O
• Tools Reference - Complete tool documentation
• Multi-Agent - Agent orchestration
• Streaming - Real-time responses
• CLI - Scaffold projects