Skip to navigation

Overview

Build, deploy, and manage AI agents with Astropods
View as Markdown

Astropods is a platform for building, deploying, and managing AI agents. It provides agent-native infrastructure for models, knowledge, tools, messaging, and observability.

With Astropods, you can develop an agent locally, publish it as a versioned blueprint, and deploy it as a hosted agent.

Watch on YouTube.

Get started

Explore Astropods

Core concepts

  • Project — A local agent codebase with an astropods.yml spec. Use ast project create to create a project, ast project configure to set credentials, and ast project start to run it locally.

  • Astropods Spec — The declarative YAML configuration in astropods.yml that describes an agent’s topology, including its container image, models, knowledge stores, tools, and ingestion pipelines. See Astropods Spec.

  • Blueprint — A versioned snapshot of a project pushed to the Astropods registry with ast blueprint push.

  • Agent — A running instance of a blueprint deployed on the platform. Deploy an agent with ast blueprint deploy and manage it with ast agent commands.

  • Account — The scope for Astropods platform resources. Blueprints, agents, and secrets belong to an account.

  • Secrets — Encrypted credentials stored in the account vault and referenced when you deploy an agent.

  • Agent Card — An AGENT.md file that controls how a blueprint appears in the catalog, including its description, tags, authors, and integrations. See Make your agent discoverable.