OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
Agent Frameworks

Agent Foundation (a13n)

Agent Foundation (a13n) is an Apache-2.0 open-source Python library and self-hosted platform for building and operating AI agent systems. It combines a Service for managed agents and APIs, a Harness for embedding and extending the runtime, and Harness UI for interactive work with models, tools, Skills and environments.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Agent Foundation (a13n) — OpenViST project profile in Agent Frameworks.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use Agent Foundation when an agent project needs more than a model call: managed execution, memory, sandboxing, configurable integrations or recovery for durable workflows. Harness UI is the documented entry point for interactive exploration before building an application.

Who it is for

  • Developers building AI agent applications
  • Teams needing a self-hosted agent runtime
  • Individuals exploring agent configurations through a terminal or browser workbench

Documented uses

  • Deploy managed AI agents in a self-hosted service
  • Add persistent memory to agent applications
  • Run computer-use work inside sandboxed environments
  • Support durable, long-running agent workflows
  • Connect models, web-data services and remote MCP servers
Before you choose

Requirements & limits

The documented Service setup runs with Docker Compose and requires Docker with Docker Compose plus either a model-provider API key or an eligible ChatGPT plan. The first launch creates the administrator account; additional users join through administrator invitations.

  • Service, Harness and Harness UI serve different roles: managed deployment, embedded runtime extension and interactive exploration.
  • The documented implementation uses Python and requires Python 3.13+ for the repository tooling, while the Compose path does not require a source checkout, Make, Python or Node.js.
  • Execution environments include Docker and listed hosted sandbox providers; the Local directory environment is available only in development.
  • The stack includes PostgreSQL and Redis and accesses the host Docker Engine for agent execution.
  • The project is in active 0.x development, so minor releases may change APIs and configuration.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Start the self-hosted Service with Docker Compose

Launch the documented Service stack and create a first agent.

Before you start

  • Docker with Docker Compose
  • A model provider API key or an eligible ChatGPT plan
  • The published Compose file

Steps

  1. Download the a13n Service Compose file and start it with Docker Compose.

  2. Open http://127.0.0.1:8080 in a browser.

  3. Register the first administrator account with an email address and a password of at least 8 characters.

  4. Under Models, add a provider using an API key or Sign in with ChatGPT, then select a model.

  5. Create an agent and send a first message with Try agent.

Expected outcome

A local Service deployment with an administrator, connected model and agent available for an initial interaction.

Guide limitations

  • The project is in active 0.x development, so APIs and configuration may change between minor releases.
  • The Compose setup mounts the host Docker socket and binds to loopback; consult the deployment guide before exposing Service to other machines.
  • The source Compose file uses the published latest image, while release assets pin a release version.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
3 October 2026
Original website title
converge-ai-labs/agent-foundation · Official homepage title · retained in its source language
How we review sourcesSuggest a correction