entra-agent-user

por github

Crea usuarios agente en Microsoft Entra ID para habilitar agentes de IA que actúen como trabajadores digitales con acceso de identidad de usuario. Aprovisiona identidades de usuario especializadas (tokens idtyp=user) vinculadas a identidades de agente, permitiendo que los agentes accedan a API exclusivas para usuarios como buzones de Exchange, Teams y organigramas. Requiere una identidad de agente principal creada a partir de un blueprint de identidad de agente; admite relación 1:1 con asignación opcional de gerente y aprovisionamiento de licencias. Incluye HTTP y PowerShell paso a paso...

npx skills add https://github.com/github/awesome-copilot --skill entra-agent-user

SKILL: Creating Agent Users in Microsoft Entra Agent ID

Overview

An agent user is a specialized user identity in Microsoft Entra ID that enables AI agents to act as digital workers. It allows agents to access APIs and services that strictly require user identities (e.g., Exchange mailboxes, Teams, org charts), while maintaining appropriate security boundaries.

Agent users receive tokens with idtyp=user, unlike regular agent identities which receive idtyp=app.


Prerequisites

  • A Microsoft Entra tenant with Agent ID capabilities
  • An agent identity (service principal of type ServiceIdentity) created from an agent identity blueprint
  • One of the following permissions:
    • AgentIdUser.ReadWrite.IdentityParentedBy (least privileged)
    • AgentIdUser.ReadWrite.All
    • User.ReadWrite.All
  • The caller must have at minimum the Agent ID Administrator role (in delegated scenarios)

Important: The identityParentId must reference a true agent identity (created via an agent identity blueprint), NOT a regular application service principal. You can verify by checking that the service principal has @odata.type: #microsoft.graph.agentIdentity and servicePrincipalType: ServiceIdentity.


Architecture

Agent Identity Blueprint (application template)
    │
    ├── Agent Identity (service principal - ServiceIdentity)
    │       │
    │       └── Agent User (user - agentUser) ← 1:1 relationship
    │
    └── Agent Identity Blueprint Principal (service principal in tenant)
ComponentTypeToken ClaimPurpose
Agent IdentityService Principalidtyp=appBackend/API operations
Agent UserUser (agentUser)idtyp=userAct as a digital worker in M365

Step 1: Verify the Agent Identity Exists

Before creating an agent user, confirm the agent identity is a proper agentIdentity type:

GET https://graph.microsoft.com/beta/servicePrincipals/{agent-identity-id}
Authorization: Bearer <token>

Verify the response contains:

{
  "@odata.type": "#microsoft.graph.agentIdentity",
  "servicePrincipalType": "ServiceIdentity",
  "agentIdentityBlueprintId": "<blueprint-id>"
}

PowerShell

Connect-MgGraph -Scopes "Application.Read.All" -TenantId "<tenant>" -UseDeviceCode -NoWelcome
Invoke-MgGraphRequest -Method GET `
  -Uri "https://graph.microsoft.com/beta/servicePrincipals/<agent-identity-id>" | ConvertTo-Json -Depth 3

Common mistake: Using an app registration's appId or a regular application service principal's id will fail. Only agent identities created from blueprints work.


Step 2: Create the Agent User

HTTP Request

POST https://graph.microsoft.com/beta/users/microsoft.graph.agentUser
Content-Type: application/json
Authorization: Bearer <token>

{
  "accountEnabled": true,
  "displayName": "My Agent User",
  "mailNickname": "my-agent-user",
  "userPrincipalName": "[email protected]",
  "identityParentId": "<agent-identity-object-id>"
}

Required Properties

PropertyTypeDescription
accountEnabledBooleantrue to enable the account
displayNameStringHuman-friendly name
mailNicknameStringMail alias (no spaces/special chars)
userPrincipalNameStringUPN — must be unique in the tenant (alias@verified-domain)
identityParentIdStringObject ID of the parent agent identity

PowerShell

Connect-MgGraph -Scopes "User.ReadWrite.All" -TenantId "<tenant>" -UseDeviceCode -NoWelcome

$body = @{
  accountEnabled    = $true
  displayName       = "My Agent User"
  mailNickname      = "my-agent-user"
  userPrincipalName = "[email protected]"
  identityParentId  = "<agent-identity-object-id>"
} | ConvertTo-Json

Invoke-MgGraphRequest -Method POST `
  -Uri "https://graph.microsoft.com/beta/users/microsoft.graph.agentUser" `
  -Body $body -ContentType "application/json" | ConvertTo-Json -Depth 3

Key Notes

  • No password — agent users cannot have passwords. They authenticate via their parent agent identity's credentials.
  • 1:1 relationship — each agent identity can have at most one agent user. Attempting to create a second returns 400 Bad Request.
  • The userPrincipalName must be unique. Don't reuse an existing user's UPN.

Step 3: Assign a Manager (Optional)

Assigning a manager allows the agent user to appear in org charts (e.g., Teams).

PUT https://graph.microsoft.com/beta/users/{agent-user-id}/manager/$ref
Content-Type: application/json
Authorization: Bearer <token>

{
  "@odata.id": "https://graph.microsoft.com/beta/users/{manager-user-id}"
}

PowerShell

$managerBody = '{"@odata.id":"https://graph.microsoft.com/beta/users/<manager-user-id>"}'
Invoke-MgGraphRequest -Method PUT `
  -Uri "https://graph.microsoft.com/beta/users/<agent-user-id>/manager/`$ref" `
  -Body $managerBody -ContentType "application/json"

Step 4: Set Usage Location and Assign Licenses (Optional)

A license is needed for the agent user to have a mailbox, Teams presence, etc. Usage location must be set first.

Set Usage Location

PATCH https://graph.microsoft.com/beta/users/{agent-user-id}
Content-Type: application/json
Authorization: Bearer <token>

{
  "usageLocation": "US"
}

List Available Licenses

GET https://graph.microsoft.com/beta/subscribedSkus?$select=skuPartNumber,skuId,consumedUnits,prepaidUnits
Authorization: Bearer <token>

Requires Organization.Read.All permission.

Assign a License

POST https://graph.microsoft.com/beta/users/{agent-user-id}/assignLicense
Content-Type: application/json
Authorization: Bearer <token>

{
  "addLicenses": [
    { "skuId": "<sku-id>" }
  ],
  "removeLicenses": []
}

PowerShell (all in one)

Connect-MgGraph -Scopes "User.ReadWrite.All","Organization.Read.All" -TenantId "<tenant>" -NoWelcome

# Set usage location
Invoke-MgGraphRequest -Method PATCH `
  -Uri "https://graph.microsoft.com/beta/users/<agent-user-id>" `
  -Body '{"usageLocation":"US"}' -ContentType "application/json"

# Assign license
$licenseBody = '{"addLicenses":[{"skuId":"<sku-id>"}],"removeLicenses":[]}'
Invoke-MgGraphRequest -Method POST `
  -Uri "https://graph.microsoft.com/beta/users/<agent-user-id>/assignLicense" `
  -Body $licenseBody -ContentType "application/json"

Tip: You can also assign licenses via the Entra admin center under Identity → Users → All users → select the agent user → Licenses and apps.


Provisioning Times

ServiceEstimated Time
Exchange mailbox5–30 minutes
Teams availability15 min – 24 hours
Org chart / People searchUp to 24–48 hours
SharePoint / OneDrive5–30 minutes
Global Address ListUp to 24 hours

Agent User Capabilities

  • ✅ Added to Microsoft Entra groups (including dynamic groups)
  • ✅ Access user-only APIs (idtyp=user tokens)
  • ✅ Own a mailbox, calendar, and contacts
  • ✅ Participate in Teams chats and channels
  • ✅ Appear in org charts and People search
  • ✅ Added to administrative units
  • ✅ Assigned licenses

Agent User Security Constraints

  • ❌ Cannot have passwords, passkeys, or interactive sign-in
  • ❌ Cannot be assigned privileged admin roles
  • ❌ Cannot be added to role-assignable groups
  • ❌ Permissions similar to guest users by default
  • ❌ Custom role assignment not available

Troubleshooting

ErrorCauseFix
Agent user IdentityParent does not existidentityParentId points to a non-existent or non-agent-identity objectVerify the ID is an agentIdentity service principal, not a regular app
400 Bad Request (identityParentId already linked)The agent identity already has an agent userEach agent identity supports only one agent user
409 Conflict on UPNThe userPrincipalName is already takenUse a unique UPN
License assignment failsUsage location not setSet usageLocation before assigning licenses

References

Más skills de github

console-rendering
github
Instrucciones para usar el sistema de renderizado en consola basado en etiquetas de struct en Go
official
acquire-codebase-knowledge
github
Usa esta habilidad cuando el usuario solicite explícitamente mapear, documentar o incorporarse a un código base existente. Actívala para indicaciones como "mapea este código base", "documenta…
official
acreadiness-assess
github
Run the AgentRC readiness assessment on the current repository and produce a static HTML dashboard at reports/index.html. Wraps `npx github:microsoft/agentrc…
official
acreadiness-generate-instructions
github
Genera archivos de instrucciones de agente de IA personalizados mediante el comando de instrucciones de AgentRC. Produce .github/copilot-instructions.md (por defecto, recomendado para Copilot en VS…)
official
acreadiness-policy
github
Ayudar al usuario a seleccionar, redactar o aplicar una política de AgentRC. Las políticas personalizan la puntuación de readiness desactivando comprobaciones irrelevantes, anulando impacto/nivel, estableciendo…
official
add-educational-comments
github
Añade comentarios educativos a archivos de código para convertirlos en recursos de aprendizaje efectivos. Adapta la profundidad y el tono de las explicaciones a tres niveles de conocimiento configurables: principiante, intermedio y avanzado. Solicita automáticamente un archivo si no se proporciona ninguno, con una lista numerada para una selección rápida. Expande los archivos hasta un 125% utilizando solo comentarios educativos (límite estricto: 400 líneas nuevas; 300 para archivos de más de 1,000 líneas). Conserva la codificación del archivo, el estilo de sangría, la corrección sintáctica y...
official
adobe-illustrator-scripting
github
Escribir, depurar y optimizar scripts de automatización de Adobe Illustrator usando ExtendScript (JavaScript/JSX). Úselo al crear o modificar scripts que manipulen…
official
agent-governance
github
Políticas declarativas, clasificación de intenciones y registros de auditoría para controlar el acceso y comportamiento de herramientas de agentes de IA. Las políticas de gobernanza componibles definen herramientas permitidas/bloqueadas, filtros de contenido, límites de velocidad y requisitos de aprobación, almacenados como configuración, no como código. La clasificación semántica de intenciones detecta indicaciones peligrosas (exfiltración de datos, escalada de privilegios, inyección de indicaciones) antes de la ejecución de herramientas mediante señales basadas en patrones. El decorador de gobernanza a nivel de herramienta aplica políticas en funciones...
official