gemini-managed-agents-api
Pass
Audited by Gen Agent Trust Hub on May 19, 2026
Risk Level: SAFE
Full Analysis
- Authentication Management: The skill instructs users to use Application Default Credentials (ADC) and the
gcloudCLI to generate access tokens. This is a standard, secure method for authenticating with Google Cloud services. - External Resource Configuration: It documents how to mount resources from Google Cloud Storage (GCS) and the Skill Registry. These operations are performed through authenticated API calls to official Google endpoints (
aiplatform.googleapis.com). - Model Context Protocol (MCP) Integration: The documentation includes examples for configuring external MCP servers, including the use of authentication headers. It notes that the platform is designed to route these headers only to the specified server, which is a key security consideration for tool integration.
- Network Environment Controls: The skill provides examples of network allowlisting (e.g., using a wildcard for domains). This is presented as a configuration option for the managed agent's environment, allowing users to define the network boundaries for their custom agents.
- Resource Management (CRUD): The core functionality involves standard Create, Read, Update, and Delete operations on agent resources via RESTful APIs, following common developer workflows for cloud infrastructure management.
Audit Metadata