SDK Quickstart

Get your first AI agent running in 5 minutes.

Prerequisites

• Sekuire CLI installed
• LLM API key (OpenAI, Anthropic, etc.)

Optional (platform mode only):

• Sekuire API key from your dashboard

---

Step 1: Create a Project

Use the CLI to scaffold a new agent:

Terminal

sekuire init --name my-agent --language python
cd my-agent

This creates a sekuire.yml config file and project structure.

---

Step 2: Configure Your Agent

Edit sekuire.yml to set your LLM provider:

sekuire.yml

project:
  name: "my-agent"
  version: "1.0.0"

agents:
  assistant:
    name: "AI Assistant"
    system_prompt: "./prompts/assistant.md"
    llm:
      provider: "openai"
      model: "gpt-4-turbo"
      api_key_env: "OPENAI_API_KEY"
      temperature: 0.7
    memory:
      type: "buffer"
      max_messages: 10

Set your API key in the environment:

export OPENAI_API_KEY="sk-..."
# or add to .env file

---

Step 3: Install the SDK

TypeScript

npm install @sekuire/sdk

Python

pip install sekuire-sdk

Rust

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

---

Step 4: Use the Agent

TypeScript

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

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

  // Chat with the agent
  const response = await agent.chat('What is the capital of France?');
  console.log(response);

  // Continue the conversation
  const followUp = await agent.chat('What is its population?');
  console.log(followUp);

  // Get conversation history
  const history = agent.getHistory();
  console.log(`Messages: ${history.length}`);
}

main();

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('What is the capital of France?')
    print(response)

    # Continue the conversation
    follow_up = await agent.chat('What is its population?')
    print(follow_up)

    # Get conversation history
    history = agent.get_history()
    print(f"Messages: {len(history)}")

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("What is the capital of France?", None).await?;
    println!("{}", response);

    // Continue the conversation
    let follow_up = agent.chat("What is its population?", None).await?;
    println!("{}", follow_up);

    // Get conversation history
    let history = agent.get_history();
    println!("Messages: {}", history.len());

    Ok(())
}

---

Step 5: Run It

TypeScript

npx tsx main.ts
# or
node --loader ts-node/esm main.ts

Python

python main.py

Rust

cargo run

tip

This quickstart runs in standalone mode by default. Add platform credentials later for managed features.

---

Next Steps

• Agent API - Deep dive into agent methods
• Built-in Tools - Add capabilities to your agent
• Streaming - Real-time token responses
• Publishing - Push to the registry