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

Agentfm Core

Agentfm Core is peer-to-peer compute infrastructure for running containerized AI agents on participating computers. It provides worker, boss and relay roles: workers run agents in Podman sandboxes, bosses orchestrate and dispatch tasks, and relays support peer discovery and NAT traversal.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Agentfm Core — OpenViST project profile in Agent Infrastructure.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use Agentfm Core when an agent can be packaged as a container and its tasks should run on participating machines in a peer-to-peer mesh, rather than on one host alone.

Who it is for

  • Developers packaging AI agents as containers for distributed execution
  • Teams that want to dispatch agent tasks across participating computers
  • Users seeking a desktop console or an OpenAI-compatible gateway for mesh workloads

Documented uses

  • Run a containerized agent in a Podman sandbox on a worker machine
  • Dispatch tasks from a desktop app, terminal interface or HTTP gateway
  • Route work using reported CPU, GPU, memory and queue information
  • Collect task output and returned artifact files
Before you choose

Requirements & limits

The README lists desktop installers for macOS and Linux; it says the Windows app is in progress. Its desktop quick start says the bundled node connects to the public mesh. The README notes that the v1 installers are unsigned and macOS may block the first launch.

  • The documented execution model packages agents as containers and uses Podman sandboxing.
  • The stack identifies libp2p for peer-to-peer networking and separates worker, boss and relay roles.
  • The desktop app bundles the mesh backend; the README also describes CLI and OpenAI-compatible HTTP gateway access.
  • The CLI example uses a local Ollama-hosted Llama 3.2 model.

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

Run a local-model worker and submit a task through the gateway

Start a Podman-backed worker using Ollama's Llama 3.2 model, then submit a chat-completions request to the local Agentfm Core gateway.

Before you start

  • macOS with Homebrew, or Ubuntu using apt for Podman installation
  • Ollama and the Llama 3.2 model
  • Agentfm Core installed from its install script or a release binary
  • An agent directory such as the README example path

Steps

  1. Install Podman; on macOS, initialize and start its machine.

  2. Install Ollama and run the Llama 3.2 model.

  3. Install Agentfm Core and start a worker with the agent directory, image name, model and agent name.

  4. In another terminal, start the API mode on port 8080 and send a chat-completions request to the local gateway.

Expected outcome

The gateway returns the worker's response; files produced in /tmp/output are described as returning in a ZIP under ./agentfm_artifacts/.zip.

Guide limitations

  • The documented sequence is an example using macOS or Ubuntu and Llama 3.2.
  • The README does not specify workload limits or supported container formats.
Documentation 1 ↗
COMPARE OPTIONS

Compare options for Agent orchestration

These projects document a common task and share a project role. This is a source-based comparison, not an execution benchmark or a guarantee of interchangeability.

ProjectDocumented purposePlatformAccess and requirements
Agentfm CoreRun a containerized agent in a Podman sandbox on a worker machineDesktop, Docker, Self-hostedThe README lists desktop installers for macOS and Linux; it says the Windows app is in progress. Its desktop quick start says the bundled node connects to the public mesh. The README notes that the v1 installers are unsigned and macOS may block the first launch.
PlanoRoute incoming requests to agents based on their declared roles.Self-hosted, ServerPlano is public source code under the Apache-2.0 license. The README describes hosted Plano models free of charge in the US-central region for a first-run developer experience; for production scale it points to running models locally or contacting the project for API keys.
TensorlakeRun an agent or its tools inside a stateful MicroVM sandbox.Cloud, ServerThe README directs cloud users to sign up for an account, obtain an API key, and authenticate with the CLI. It documents installation of the Python SDK with pip and distribution of the `tl` CLI as a standalone binary. The repository is public and licensed under Apache-2.0.
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
15 October 2026
Original website title
Agent-FM/agentfm-core · Official homepage title · retained in its source language
How we review sourcesSuggest a correction