Now in Early Access

The Operating System for
Agentic Engineering.

Orchestrate ecosystems, runs, releases, and AI workers across local and Kubernetes-backed environments—with live logs, telemetry, and one control plane.

$amof run --ecosystem my-stack --target k8s
AMOF Control Plane
api-gateway
worker-agent
auth-svc
web-ui
Release
v2.1.0-rc.1
run-0xf3a · active · 3 agents · 2 repos
12s elapsed
00:02[master] ecosystem loaded — 3 services, 2 repos
00:04[worker-1] cloned api-gateway @ release/v2.1
00:07[worker-2] running integration tests on auth-svc
00:10[master] promoting v2.1.0-rc.1 → stable
promote auth-svc to stable and tail the deploy logs

The Problem

Modern agentic development has no operating layer.

AI agents are powerful in isolation. But the moment you run multiple agents across multiple repos in real environments, the developer experience collapses. You become the glue.

Terminal & Context Hell

12 tabs, 5 editors, 3 browsers. Every context switch kills your flow state.

Multi-Repo Dependency Nightmare

One feature, five repos. Branches, configs, and deps out of sync across every service.

Broken Localhost (Port Collisions)

Services fighting over ports and leaking session state. Zero isolation between projects.

Blind Agents, Zero Observability

AI workers running in the dark. No run history, no telemetry, no idea what they changed or why they broke.

No Single Pane of Glass

Stitching logs across terminals and dashboards to figure out what ran, what failed, and when.

Release Workflow Chaos

Versioning lives in scripts and team memory. No unified path from alpha to rc to stable.

Why AMOF

The Solution

A coherent system. Not another patchwork of tools.

AMOF replaces the patchwork of terminals, scripts, and dashboards with a single control plane where ecosystems, runs, agents, and releases are all connected, visible, and operable from one surface.

Your Entire Stack, One Ecosystem

Define your entire multi-repo stack as a single ecosystem. AMOF coordinates clones, branches, configs, and execution order across every service.

Every Operation is a Tracked Run

Every action is a run with a timeline, a log stream, and a result. No more guessing what happened. Runs are first-class, observable, and auditable.

Chat to Your Control Plane

Issue commands through a visual chat interface backed by MCP tools. Promote a release, tail a deploy, or query run state without leaving the app.

Local Laptop to Kubernetes Pod

Start with local processes. Promote to WSL, devcontainers, or fully isolated Kubernetes pod runners with Helm charts. Same control plane, any target.

See Every Agent as It Thinks

Stream structured output from every agent in real time. See the full master/worker hierarchy, token usage, task completion rates, and run health at a glance.

Release Lifecycle, Built In

Manage the full release lifecycle from alpha to rc to stable. SemVer versioning, validation gates, one-click promotion, and audit trail built into the control plane.

How It Works

From ecosystem config to live execution in four steps.

01

Define Your Ecosystem

Connect your repos, declare services, and set the execution target. AMOF generates a workspace model that understands the shape of your stack.

>amof ecosystem init --repos api,auth,web
02

Launch Runs & Promote Releases

Install dependencies, deploy services, execute agent tasks, or promote a release from alpha to stable. Every operation is a tracked run launched from the control plane or the chat surface.

>amof run --task deploy --target local
03

Observe Everything

Stream live logs, inspect run history, view telemetry dashboards, and trace the full master/worker agent hierarchy. See who did what, when, and why.

>live stream in control plane
04

Execute Anywhere

Run locally during development. Switch to WSL or devcontainers for isolation. Graduate to Kubernetes-backed pod runners with Helm for production-grade workloads.

>amof run --target k8s --namespace staging

Core Capabilities

Everything you need to operate agentic software.

Not a feature checklist. A connected system where orchestration, execution, observability, and release management work together.

Orchestrate

Multi-Repo Stack Orchestration

Define your entire multi-repo stack as a single ecosystem. Coordinate services, configs, and execution order from one manifest.

Synchronized Branch & Build

Branch, build, and synchronize across repos in lockstep. One operation fans out to every service in the ecosystem.

Execute

Instant Isolated Dev Environments

Spin up fully isolated execution contexts in seconds. No more port collisions, leaked state, or manual Docker wrangling.

Key Benefit

Local to Kubernetes in One Flag

Start locally, graduate to WSL, devcontainers, or Kubernetes pods via Helm charts. Same control plane, any execution target.

Observe

Watch Agents Think in Real-Time

Stream SSE logs from every running agent. See reasoning steps, tool calls, and outputs as they happen -- not after the run is over.

Key Benefit

Telemetry That Makes Agents Measurable

Track token usage, task completion rates, agent uptime, and error frequency across every run with structured dashboards.

Control

Visual Agent Chat & MCP Control

Issue structured commands through a conversational UI backed by MCP tools. Promote releases, query runs, or tail logs from one chat.

Full Run Timeline & History

Every operation is a tracked run with inputs, outputs, agent activity, and outcome. Query past runs for debugging and compliance.

Release

Built-In Release Management

Manage the full semver lifecycle inside the control plane. Track alpha, rc, and stable states. Validate before every promotion.

One-Click Release Promotion

Promote releases from alpha to rc to stable with a single action. Validation gates and automatic rollback paths built in.

Release Management

Releases are first-class. Not an afterthought.

AMOF makes versioning, promotion, and release state visible inside the same control plane where you orchestrate and observe. No more stitching together CI scripts, manual tagging, and Slack messages to coordinate a release across five services.

alpharcstable
api-gatewayv3.2.0
auth-svcv2.1.0-rc.2
worker-agentv1.0.0-alpha.4

SemVer Versioning

Every release follows semantic versioning. AMOF tracks major, minor, and patch versions across your ecosystem services.

Lifecycle Promotion

Move releases through alpha, release candidate, and stable stages. Each stage is a tracked state transition, not an ad-hoc script.

Validation Before Promotion

Define validation gates that must pass before a release can be promoted. Tests, checks, and approvals happen inside the control plane.

Audit Trail & Release Record

Every promotion, rollback, and release event is recorded. Query the full release history for any service in your ecosystem.

Pricing

Lifetime access. Pay once, build forever.

Start free with the community CLI. Upgrade to the full control plane when you are ready. No subscriptions, no recurring fees.

Community

For individuals exploring AMOF.

Free
Join Waitlist
  • Open-source CLI (coming soon)
  • Local runner support
  • Bring your own LLM keys
  • Community Discord access
  • Public documentation
Limited to 50 Seats

Solo Developer

For serious solo builders and indie hackers.

$499$249 one-time
Buy Solo License
  • Lifetime access -- pay once, own it forever
  • Everything in Community
  • Full Control Plane UI (desktop & web)
  • Visual Agent Chat & MCP control surface
  • Real-time SSE logs & telemetry dashboards
  • Complete run history & execution timeline
  • Instant isolated dev environments (K3d)
  • Release workflow visibility & state tracking
  • Every future update included

Pro / Startup

For small teams and production workloads.

$599 one-time
Buy Pro License
  • Everything in Solo Developer
  • Kubernetes Helm Charts for self-hosting
  • Isolated pod runners for agents
  • Release promotion workflows
  • Multi-agent orchestration support
  • Priority email support from the founder

Enterprise

Need SSO, RBAC, SLA-backed support, or custom integrations?

Book a Demo

All purchases are subject to our Terms & Conditions and Privacy Policy. Payments are processed securely via Stripe.

FAQ

Under the Hood

Technical details for builders who want to know what they are getting.

Product & Licensing

AI Orchestration

Infrastructure & Isolation

Git & State Management

Control Plane & UI

Security & Observability

Extensibility & Developer Experience

Your agents deserve a real operating layer.

Stop being the glue between repos, terminals, and disconnected AI workers. AMOF gives you one control plane for orchestration, isolated execution, live observability, and release management. Built for engineers who ship.