Open source · Apache-2.0 · Pre-alpha

OpenEchelon
the operating system for AI organizations

A model-agnostic runtime for structured AI organizations — hierarchical multi-agent systems where persistent AI employees operate within defined authority, permission, memory, and resource boundaries, and report ultimately to a human owner.

Status: pre-alpha Apache-2.0 Python 3.12+ Local-first Provider-independent

You talk to the CEO. The organization handles the rest.

Complexity belongs inside the organization, not in the user's lap.

Most multi-agent systems expose their complexity directly. You choose agents, wire workflows, select models, coordinate handoffs, and decide which agent does what. OpenEchelon models an actual organization instead: executives, directors, managers, specialists, and workers, each with defined responsibilities, authority, tools, memory, budgets, and reporting lines.

                         YOU
                          |
                          v
                         CEO
                          |
            +-------------+-------------+
            |             |             |
            v             v             v
           CTO           CRO           CQO
            |             |             |
            v             v             v
        Directors     Directors     Directors
            |             |             |
            v             v             v
         Managers      Managers      Managers
            |             |             |
            v             v             v
         Workers       Workers       Workers

The CEO receives the objective, delegates through the organization, monitors progress, resolves conflicts, commissions reviews, manages scarce model resources, and returns the consolidated result.

Hierarchy, not swarm

The difference is operational, not cosmetic.

Flat multi-agent swarmOpenEchelon
StructureLead plus peer agentsMulti-level organization with departments
ContextOften one shared contextMinimum sufficient context per role
Model choiceUsually one provider for all agentsPer-task assignment by a Resource Governor
ReasoningFrequently maximized everywhereBudgeted by rank and task requirement
IdentityAgent equals prompt plus modelPersistent employee with history and performance record
ScalingMore agents means more running modelsLogical identity scales separately from execution
EscalationRetry the same agentStructured escalation to a stronger resource
AuditConversation historyTraceable task tree with artifacts and reviews

In a flat swarm, ten agents on one problem tends to mean ten expensive contexts reasoning about the same thing. Orchestrator-worker systems buy their quality with roughly an order of magnitude more tokens than a single agent, and that multiplier compounds through every handoff. OpenEchelon treats that as the problem to solve, not the price of admission.

What holds it together

Forty architectural constraints. Six that decide the shape.

PRINCIPLE 2

Human authority is final

The organization never structurally places itself above its owner. Autonomy exists within delegated authority; it is not sovereignty.

PRINCIPLE 4

Agents are not models

An employee is a persistent identity with a role, manager, permissions, memory, and performance history. Execution resources are assigned per task.

PRINCIPLE 5

Identity scales past execution

A thousand registered employees might mean twenty actively executing. One loaded local model can serve hundreds of logical workers over time.

PRINCIPLE 8

Cheapest sufficient intelligence

Work starts on local or inexpensive models and escalates only when the attempt proves insufficient — because it is justified, not because a powerful model happens to be available.

PRINCIPLE 14

Managers compress context

Raw worker output does not propagate upward unchanged. Reports keep references to underlying evidence so deeper inspection stays possible.

PRINCIPLE 35

Security is architectural

Prompt injection, authority escalation, information boundaries, credential isolation, and approval gates are design concerns — not prompt instructions added after autonomy.

Where the project actually is

Pre-alpha, and specific about it.

The Phase 0 organizational data model exists and is tested: employees, departments, grants, tasks and delegation trees, typed messages, artifacts, reviews, execution resources, quota, and routing decisions. A Resource Governor spike measures the escalation ladder against labelled scenarios. There is no organization runtime and no released package.

The spike is deliberately sequenced first. The whole cost argument rests on a component that decides whether a cheap answer was good enough, and published routing work suggests that component generalizes poorly across task classes. Building an organization runtime on an unvalidated estimator would spend the larger investment before testing the premise.

PhaseContentsState
Phase 0Organizational data model, identity, permissions, task graph, message protocol, provider abstraction, artifactsBuilt and tested
Phase 0Resource GovernorSpike, measured
Phase 1Minimum viable organization: CEO, three managers, worker pool, local execution, CLI agent integration, review loopNot started
Phase 2Departments, dynamic hiring, persistent memory, approval workflows, performance trackingNot started
Phases 3–4Ecosystem interoperability, large organizations, learned routingVision

Frequently asked questions

Direct answers. The full set is in the repository.

What is OpenEchelon?

OpenEchelon is an open-source, model-agnostic runtime for structured AI organizations. It models a hierarchy of AI agents — CEO, executives, directors, managers, specialists, and workers — where each agent has a defined role, manager, permissions, memory, reasoning budget, and reporting line. The human owner gives an objective to the CEO and retains final authority over everything the organization does.

How is this different from a multi-agent swarm?

A swarm is flat: one lead agent fans work out to peer agents, usually sharing one context and one model. OpenEchelon is hierarchical, and the hierarchy carries operational meaning — information access, permissions, reasoning budgets, escalation paths, and model tier all vary by rank. In a flat swarm, ten agents on one problem tends to mean ten expensive contexts reasoning about the same thing. Here, a worker gets a bounded assignment on a cheap model, and only the levels that need judgement pay for judgement.

Does a thousand agents mean a thousand running models?

No. Logical identity is deliberately separated from execution. A logical employee exists persistently without a dedicated model process, container, GPU allocation, or running session. A thousand registered employees might mean fifty assigned and twenty actively executing against a limited pool, and one loaded local model can serve hundreds of logical workers over time.

Which models and providers does it require?

None. Local inference (Ollama, llama.cpp, OpenAI-compatible servers), CLI coding agents (Codex CLI, Claude Code, Gemini and Antigravity), and hosted APIs all sit behind adapters. OpenEchelon is designed to run meaningfully in a local-only configuration. Cloud and frontier resources enhance capability but must not define the architecture.

How does it decide which model runs a task?

A Resource Governor assigns execution resources per task, starting with the cheapest intelligence that could plausibly succeed and escalating when a sufficiency estimator judges the answer inadequate. Inputs include task complexity and importance, uncertainty and risk, remaining provider quota and reset windows, latency, local hardware availability, historical success rates, cost, concurrency, and organizational rank.

Reasoning follows the same discipline. Rank sets a ceiling on spending, never a cap on what a task needs — a task requiring more reasoning than its role permits escalates rather than running under-powered.

What stops an agent doing something dangerous?

Architectural constraints rather than a single guardrail prompt. Permissions are organizational and enforced technically where possible. Information access follows organizational need, so a worker does not receive the whole context. Sensitive external actions — sending messages, publishing, deleting data, financial transactions, production deployments — require an approval chain whose height depends on risk, up to human approval. Execution is interruptible: work can be paused, cancelled, or have its resource access revoked without waiting for an autonomous process to finish voluntarily.

Is it a workflow builder or a chatbot wrapper?

Neither, by explicit non-goal. A visual workflow editor may eventually exist as an advanced feature but must never be required for normal operation; execution plans are constructed from organizational structure, policy, and available resources. And while the default interface is a conversation with the CEO, the substance behind it is persistent identities, delegation trees, permission boundaries, artifact provenance, independent review, and resource governance.

Is OpenEchelon usable today?

Not yet. It is pre-alpha. The Phase 0 organizational data model and a Resource Governor spike exist and are tested; there is no organization runtime and no released package. The most useful contributions right now are architectural critique, prior-art analysis, and concrete proposals for the core data models and protocols.