Quick Start

Get AgentOven up and running in under five minutes.

Get AgentOven up and running in under five minutes. You need Rust 1.86+ and Docker installed on your machine.

1. Install the CLI

The AgentOven CLI is distributed as a Rust crate. You can also install via pip or npm if you prefer.

terminal

$ cargo install agentoven-cli
✓ Installed agentoven v0.1.0

2. Create a Kitchen (workspace)

A Kitchen is an isolated workspace that holds your agents, recipes, configs, and secrets. Think of it as a Kubernetes namespace for agents.

terminal

$ agentoven init my-kitchen
🏺 Created agentoven.toml
📝 Created prompts/system.md
✓ Kitchen 'my-kitchen' is ready

3. Start the Control Plane

The control plane is a Go binary backed by PostgreSQL and Redis. The easiest way to run it locally is via Docker Compose:

terminal

$ cd infrastructure && docker compose up -d
✓ postgres   — running on :5432
✓ redis      — running on :6379
✓ agentoven  — running on :8080

4. Register your first Agent

Agents are registered with the control plane along with their framework, skills, and metadata. The registry makes them discoverable via A2A.

terminal

$ agentoven agent register \
    --name research-bot \
    --framework langchain \
    --skill web-search \
    --skill summarization
✓ Agent 'research-bot' registered in kitchen 'my-kitchen'

5. Bake (deploy) the Agent

Baking validates the agent config, runs health checks, and launches the agent into the runtime with its A2A endpoint.

terminal

$ agentoven agent bake research-bot
🔥 Baking agent 'research-bot'... done!
✓ Agent is hot and ready at /agents/research-bot/a2a

6. Send a Task

Now that your agent is live, send it a task via A2A:

terminal

$ agentoven task send research-bot "Summarize the latest AI safety papers"
📨 Task submitted — id: task-a1b2c3
✓ Status: completed
✓ Result: "Here is a summary of recent AI safety research..."

7. Check Agent Status

terminal

$ agentoven agent status research-bot
┌─────────────────────────────────────────┐
│ Agent: research-bot                     │
│ Status: 🟢 hot                          │
│ Framework: langchain                    │
│ Tasks completed: 1                      │
│ Uptime: 3m 42s                          │
│ A2A endpoint: /agents/research-bot/a2a  │
└─────────────────────────────────────────┘

That's it! Your agent is now live, discoverable via A2A, and fully observable through the control plane. Next, explore the Core Concepts or dive into the CLI Reference.

Next Steps

Build your first workflow: Recipes Guide
Issue shareable access: Scoped API Keys
Learn the command surface: CLI Reference
Review the platform model: Architecture