Documentation Menu

Quickstart

Go from zero to a deployed AI agent in 10 minutes. This guide walks you through installing the CLI, creating an agent, and deploying it to a live endpoint.

Prerequisites

Install the CLI

pip install agentlift-cli

Authenticate

agentlift login

This opens a browser window for Google OAuth or email/password authentication.

Scaffold a new agent

agentlift init my-agent --template hello-world
cd my-agent

This creates a directory with an agent.yaml config file and a system prompt template.

Review the agent config

The generated agent.yaml defines your agent:

name: my-agent
version: 0.1.0

models:
  primary: claude-sonnet-4-5
  fallback:
    - gpt-4o
  temperature: 0.3
  max_tokens: 4096

tools:
  - name: create-ticket
    type: http
    endpoint: https://api.linear.app/issues
    auth: env:LINEAR_API_KEY

prompts:
  system: ./prompts/system.md

evals:
  - name: helpfulness
    type: llm-judge
    model: claude-sonnet-4-5
    criteria: "Is the response helpful and relevant?"

Push and deploy

Push your agent config to AgentLift, then deploy it:

agentlift push --team <your-team>
agentlift deploy --team <your-team>

Your agent is now live at <agentId>.agents.agentlift.dev.

Test your agent

Send a request to your agent’s OpenAI-compatible endpoint:

curl -X POST https://<agentId>.agents.agentlift.dev/v1/chat/completions \
  -H "Authorization: Bearer ag_live_..." \
  -H "Content-Type: application/json" \
  -d '{"messages": [{"role": "user", "content": "Hello!"}]}'

Or use any OpenAI-compatible SDK:

from openai import OpenAI

client = OpenAI(
    base_url="https://<agentId>.agents.agentlift.dev/v1",
    api_key="ag_live_...",
)

response = client.chat.completions.create(
    model="agent",
    messages=[{"role": "user", "content": "Hello!"}],
)

Web UI alternative

You can also create and deploy agents through the AgentLift Dashboard without using the CLI. The dashboard provides a visual editor for agent config, deployment management, and trace exploration.

Next steps