Policies and access control for AI agents

Metorial gives you granular, role-based access control over users, agents, and admins, with policies across skills, integrations, and agents, built on your existing identity.

platform.metorial.com
A screenshot showcasing the Policies + Access Control feature of Metorial.

Policies + Access Control

Control what every user and agent can do

Giving teams AI without losing control comes down to access: who can use what, and what each agent can do. Metorial provides granular, role-based access across users, agents, and admins, with policies that apply to skills, integrations, and agents, all built on the identity you already manage.

  1. 01

    Set access at the right level.

    Give each user, agent, and admin exactly the permissions they should have, and nothing more.

  2. 02

    Apply policies everywhere.

    One policy layer covers skills, integrations, and agents. Control is consistent platform-wide.

  3. 03

    Build on your identity.

    Role-based access, SAML, and group import map permissions to the roles you already manage.

What policies govern

One control layer across the platform

Access control reaches everything agents touch.

Resources

Skills and integrations

Apply policies to skills, providers, and individual tools, down to specific actions.
Agents

Per-agent control

Combine with agent identification and delegation to scope what each agent can do.
Identity

Roles and groups

Use SAML and role-based access so permissions reflect your identity provider.

What access control covers

Granular control across the board

  1. Controlled actorsUsers, agents, admins
  2. Access modelRole-based
  3. Policies apply toSkills, integrations, agents
  4. SAML and group importSupported

Set access by role. Apply policies everywhere. Scope every agent. Map it to your identity.

FAQ

Answers to common questions about policies + access control and how they fit into governed AI agent infrastructure.

  1. What can I control?
    Access for users, agents, and admins, with policies across skills, integrations, and agents. You decide who and what can do what.
  2. Yes. Access is role-based, and SAML with group import maps permissions to the roles and groups in your identity provider.
  3. Yes. With agent identification and delegation, policies scope access per agent and per session.
  4. Yes. Apply policies down to individual tools. An agent can use a provider while specific actions stay blocked.
  5. Yes. It integrates with SAML and SSO for providers. Access follows the identity people sign in with.

Give teams AI without giving up control

See how Metorial's policies and access control let you decide exactly what every user and agent can do. Book a demo, or get started.