Skip to content

Read the doctrine  ·  Install the governance plane  ·  Learn with Prof Rod

Open source Agent native Human authority Receipt driven


Organizations where agents do the work and humans own the outcomes

Zero Employee is the open-source doctrine, community, and infrastructure for building Zero Employee Organizations (ZEOs): organizations in which one human can direct fleets of AI agents while retaining authority, accountability, and ownership of the result.

The name is not a promise to remove people. It names a design problem:

If agents perform the work, how do we define scope, assign authority, control execution, account for cost, verify results, and preserve human responsibility?

This organization builds the public infrastructure needed to answer that question with working systems rather than slogans.

Start here

Understand Govern Execute Learn
Read the doctrine
Why ZEOs require a different operating model.
Install zero-employee
Scope work, enforce policy, and inspect organizational state.
Explore sovereign-agent
Run owned, inspectable agent workloads.
Visit Prof Rod
Courses, investigations, and reproducible explanations.
pip install zero-employee

The open-source stack

Project Responsibility
zero-employee · PyPI Governance plane. Schema-validated Statements of Work, deterministic policy checks, organizational state, fleet boards, cost accounting, and receipts.
sovereign-agent · PyPI Execution plane. Owned agent runtimes with filesystem-backed sessions, append-only audit trails, verifiable manifests, isolation, and recoverable work.
zeocore · PyPI Capability plane. Typed tool contracts, plugin discovery, and runner-independent production capabilities.
ffmpeg-zeo · PyPI Media capability. Typed FFmpeg filter graphs for Python, command-line systems, and coding agents.

The governing boundary

Zero Employee governs. Sovereign Agent executes. Runtime capacity never creates organizational authority.

flowchart LR
    H["Human principal"] -->|intent and authority| G["Zero Employee governance"]
    G -->|bounded request| E["Sovereign Agent execution"]
    E -->|evidence and receipts| G
    G -->|decision surface| H
Loading

The governance plane decides what may happen. The execution plane determines how permitted work runs. Evidence returns across the same boundary; authority does not travel in the opposite direction.

Operating doctrine

1. Authority must be explicit

An agent's technical ability to perform an action does not grant organizational permission to perform it.

2. Work must be bounded

Every meaningful delegation needs a defined objective, inputs, constraints, acceptance criteria, budget, and escalation path.

3. Governance must be deterministic

Prompts are advisory. Organizational constraints belong in schemas, validators, capability manifests, policy engines, and permission boundaries.

4. Execution must be inspectable

Sessions, requests, tool calls, artifacts, approvals, failures, and results must survive the process that produced them.

5. Completion requires evidence

“The agent ran” is not the same as “the work is correct.” Claims of completion require receipts that can be independently checked.

6. Humans retain responsibility

Agents can receive bounded authority. They cannot absorb the operator's accountability for irreversible consequences.

What belongs here

A project belongs in zeroemployeeorg when it provides reusable:

  • organizational doctrine or schemas
  • governance, validation, or accounting infrastructure
  • agent execution infrastructure
  • typed operational capabilities
  • reference implementations for building or operating a ZEO

Something does not belong here merely because Rod or Prof Rod uses it. Educational artifacts live under profrodai; personal work lives with Rod Rivera.

Contributing

We welcome implementation improvements, tests, documentation, reference integrations, and architectural criticism.

Before contributing:

  1. Read the target repository's CONTRIBUTING.md and architecture documentation.
  2. Open an issue for changes that alter public contracts or governance semantics.
  3. Keep policy deterministic and execution evidence inspectable.
  4. Include tests and a clear receipt showing what changed and how it was verified.

One ecosystem, three clear roles

Surface Role
Rod Rivera The human founder, researcher, and operator
Zero Employee The doctrine, community, and open-source infrastructure
Prof Rod The educational and animated surface that teaches the work

Rod originates and operates. Zero Employee codifies and ships. Prof Rod teaches and dramatizes.


Organizations where agents do the work and humans own the outcomes.

Website  ·  Governance  ·  PyPI  ·  Learn

Founded by Rod Rivera.

Popular repositories Loading

  1. sovereign-agent-lab sovereign-agent-lab Public

    Build your Sovereign Agent across 5 weeks: LangGraph automator or Rasa digital employee. Each week's assignment extends the same codebase: tools → planning → memory → production. Week 1 starts here…

    Python 10 167

  2. zeocore zeocore Public

    A typed capability-authoring framework for Python — write tools that validate input, do their work, and return a structured result.

    Python 3

  3. homework-pub-booking homework-pub-booking Public

    Python 1 98

  4. quackvideo quackvideo Public

    🦆 QuackVideo: A modern, high-performance Python video editing library built on FFmpeg. Simplify video manipulation with an intuitive, Pythonic API, designed for speed and ease of use.

    Python

  5. zeotool zeotool Public

    A QuackVerse tool built on top of QuackCore.

    Python 4

  6. agency-data-onboarding-kit agency-data-onboarding-kit Public

    Turn messy client data into clean, actionable intelligence in under 30 minutes

    Jupyter Notebook

Repositories

Showing 10 of 14 repositories

People

This organization has no public members. You must be a member to see who’s a part of this organization.

Top languages

Loading…

Most used topics

Loading…