Descope just updated its no-code identity platform, making it simpler for developers and security teams to manage how AI agents and MCP servers are authorized, control their access, handle their credentials, and set up security policies.
Descope recently unveiled Agentic Identity Hub 2.0, a significant enhancement to their popular no-code identity platform. This update is specifically designed to help manage the identities of AI agents and Model Context Protocol (MCP) servers. With this new release, developers and security teams gain a dedicated, user-friendly interface and a centralized control panel, making it much easier to handle everything from authorization and access control to credentials and security policies for these crucial AI components.
Launched on January 26, Agentic Identity Hub 2.0 empowers both MCP developers and AI agent creators to treat AI agents just like human users within the system – as ‘first-class identities.’ This means you can now manage their access with the same rigor. The update also brings OAuth 2.1 and specific tool-level scopes to both internal and external MCP servers, giving you finer control. What’s more, organizations can now enforce robust, enterprise-grade policies to perfectly govern how agents access MCP servers. Ultimately, Descope’s no-code platform aims to simplify the process of creating and adjusting identity experiences for everyone involved – your customers, partners, AI agents, and MCP servers – all through intuitive visual workflows.
Specific capabilities of Agentic Identity Hub 2.0 include:
- Agentic Identity Management: Get a clear, centralized overview of all your AI agent identities, seeing exactly how they connect to your applications, APIs, and MCP servers.
- MCP Authorization: If you’re an MCP server developer, you can now easily add secure, protocol-compliant access controls to both your internal and external MCP servers.
- Credential Vault: A secure place to store and manage all the credentials (like OAuth tokens and API keys) that your agents need to interact with other systems.
- Granular Policy Controls: Define precise rules and permissions, ensuring that only specific AI agents can access certain MCP servers and invoke particular tool-level functions.
- AI Agent Logging and Auditing: Gain full transparency into every single action your AI agents take, with detailed logs and auditing capabilities.
