OpenAI Managed Agents Deploy Agents Integration

OpenAI's agent platform, built on the OpenAI Agents SDK, gives developers OpenAI-managed capabilities for building and running agents with tools, handoffs, guardrails, and tracing.

Start for free

OpenAI Managed Agents Deploy Agents

Benefits with Ory + OpenAI Managed Agents Deploy Agents Integration

Ory integrates with OpenAI's managed agent capabilities to give teams a verified protection boundary for AI agents in production. Teams keep OpenAI-managed agent capabilities while routing gated tool calls through a local runner they control, secured with Ory Agent Security. Every gated tool call is checked against fine-grained authorization to permit or deny the action, and all activity is logged for a complete audit trail. Teams can start in Observe mode to see how agents use their tools, then switch to Enforce mode so agents act only within their authorized permissions.

The fastest way to get started with Ory Agent Security is to sign up for the Ory Network. Install the ory-agent CLI, then create an Agent Security deployment for OpenAI Managed Agents from the Ory Console or with the CLI from your agent's project directory.

brew install ory/tap/ory-agent

View OpenAI Managed Agents docs

Use Cases

— 03 use cases

Agent IAM

Secures agents running on OpenAI-managed agents with a runtime identity, fine-grained authorization on every gated tool call, and complete audit logging.

Explore Agent IAM

CIAM

Lets teams ship customer-facing agents on OpenAI-managed agents that act on behalf of end users with Ory-managed authentication and consent, following Ory best practices.

Explore CIAM

B2B IAM

Applies fine-grained authorization to every action agents take for business customers on OpenAI-managed agents, while sensitive tools stay on customer infrastructure, and scales with business growth.

Explore B2B IAM

Core Functionality

— 06 capabilities

Identity
Verifies the local runner with an Ory Agent Security runtime identity.
Authorization
Enforces fine-grained, policy-based access control for every gated tool call an agent makes on OpenAI managed agents.
Secrets
Keeps Agent Security runtime values on the customer's local runner, outside the provider-managed environment.
Isolation
Routes gated tool calls from the OpenAI-managed agent to the local runner, where Agent Security checks them.
Audit
Logs complete audit trails for agent sessions, tool invocations, and authorization decisions.
Rollout
Creates the deployment with Ory Deploy or the ory-agent CLI, ships the runtime package, and verifies the protection boundary, with Observe mode for monitoring and Enforce mode for strict permission enforcement.