GitHub Project Aims to Replace Corporate Roles with AI Permissions

Overview

A new open-source initiative called NoRoles is challenging the traditional hierarchical structures that define modern corporate environments. The project proposes a shift from job titles and managerial approval chains to a system based entirely on granular permissions and mandates. By decoupling authority from identity, the framework aims to streamline operations for organizations increasingly integrating human workers with autonomous AI agents.

The core premise is that current company structures are bottlenecked by legacy concepts like titles and middle management. In this model, work waits for approvals rather than being executed by those closest to the task. NoRoles seeks to replace the org chart with a single, open list of who can authorize irreversible actions, such as spending money, signing contracts, or sharing personal data.

"Agents now write, build, sell and pay. Companies still run on titles, managers and approvals, so the work waits."

This approach suggests that once permission is granted for specific high-impact actions, all other operational tasks should be freely executable by any authorized entity, whether human or algorithmic. The project has garnered initial attention in developer communities for its radical simplification of corporate governance logic.

The Permission-Based Model

The foundational theory behind NoRoles is detailed in its repository, which outlines a system where authority is defined by what an actor can do, not by who they are. The project distinguishes between actions that are reversible and those that are not. For irreversible actions, a specific mandate must be approved by designated holders. Once approved, the action can proceed without further managerial oversight.

Under this framework, an AI agent acts with exactly the permissions of its person, never more. This ensures that automated systems cannot exceed their authorized scope, addressing a primary concern in enterprise AI adoption: governance and risk control. The system relies on a concept called the orchestrator, which enforces a set of eight laws designed to prevent unauthorized access and ensure auditability.

Core Principles

The project’s MANIFESTO.md outlines eighteen principles guiding its development. These principles emphasize minimalism, transparency, and the elimination of redundant approval layers. The goal is to create a environment where clarity of authority replaces complexity of hierarchy.

Key aspects of the model include:

Mandates: Specific instructions or projects that require initial approval.
Permissions: Granular rights assigned to actors for specific actions.
Laws: Eight fundamental rules that the orchestrator must enforce to maintain security.

This structure allows for a more fluid organization where roles are dynamic. An individual or agent might hold permissions for marketing spend in one project but none in another, eliminating the need for broad job descriptions that often dilute accountability.

Technical Implementation

From a technical standpoint, NoRoles provides a command-line interface (CLI) built on npm. The tooling is designed to integrate directly into developer workflows, allowing teams to initialize permission structures for new ventures or existing companies. The repository includes a SPEC.md that defines the technical model, ensuring that the theoretical framework is translated into enforceable code logic.

The CLI facilitates the entire lifecycle of a permissioned action. Users can initiate a project, request permissions for agents, and approve mandates through a terminal-based interface. This integration aims to make permission management as seamless as version control or package management, tools that developers already use daily.

Workflow Example

The project documentation illustrates a typical workflow for launching a web project:

Open a Mandate: A user initiates a project, such as noroles open first-website.
Request Permissions: An agent requests permission to perform a specific action, such as spending money to buy a domain.
Approval: A human holder reviews the summary and signs off using noroles yes.
Execution: The agent runs the approved command using its own keys, ensuring it cannot exceed its mandate.

noroles ask --as me-agent --mandate first-website --permission money.spend

This process ensures that every tool call made by an AI agent goes through the NoRoles system. The project supports integration with major AI models, allowing agents to operate within the constraints defined by the human holders. This creates a closed loop of accountability where AI execution is always bound by human-approved parameters.

Implications for Enterprise AI

The rise of autonomous AI agents in business has outpaced the development of governance frameworks. Traditional corporate structures, built for human employees with fixed roles, struggle to accommodate agents that can operate across multiple functions simultaneously. NoRoles addresses this gap by providing a structural template that is role-agnostic.

By focusing on permissions rather than roles, the system allows companies to deploy AI agents without restructuring their entire human hierarchy. This could lead to significant efficiency gains, as agents can execute tasks immediately upon mandate approval, bypassing the slow chains of command typical in large organizations.

However, the model also raises questions about responsibility and legal liability. If an agent spends money or signs a contract under a approved mandate, the human holder who signed off remains the point of accountability. The system is designed to make this chain of custody clear and auditable, which is essential for regulatory compliance in sectors like finance and healthcare.

The project is currently in an early stage, with only one star and zero forks recorded in its initial repository activity. Despite this, its conceptual framework has resonated with developers interested in the intersection of organizational theory and software engineering. The MIT license allows for broad adoption and modification, encouraging community-driven evolution of the standards.

Community and Development

Development of NoRoles is guided by strict contribution standards. Every change to how approvals work requires a corresponding test in the repository's test suite. This ensures that the eight fundamental laws are never compromised by new features or refactoring. The project maintains a public roadmap and welcomes issues and pull requests from the community.

As AI capabilities continue to advance, initiatives like NoRoles may become critical infrastructure for the next generation of digital enterprises. They represent a shift toward operational agility, where the speed of decision-making is no longer limited by the rigidity of organizational charts. Whether this permission-first model will supplant traditional management structures or coexist alongside them remains an open question for the future of work.