{
  "id": 20490283,
  "title": "Govern and discover records with common metadata in AWS Agent Registry",
  "url": "https://urgent.news/2026/11/17/govern-and-discover-records-with-common-metadata-in-aws-agent-registry",
  "topic": "ai",
  "section": "AI",
  "published": "2026-11-17T20:31:04.000Z",
  "source": {
    "name": "AWS Machine Learning",
    "slug": "aws-machine-learning",
    "url": "https://aws.amazon.com/blogs/machine-learning/govern-and-discover-records-with-common-metadata-in-aws-agent-registry/"
  },
  "original_language": "en",
  "account": "In order to efficiently manage and locate growing numbers of agents, Model Context Protocol (MCP) servers, and skills, AWS Agent Registry offers a centralized registry for registering, approving, and discovering these resources. Each registry entry includes key details like name, description, protocol descriptor, and approval status.\n\nOrganizations often have additional attributes they want to track, such as business units, resource owners, risk classifications, and development environments. To support these organization-specific facts, AWS Agent Registry now allows the creation of custom metadata schemas. These schemas are composed of key-value pairs that can be applied to all registry records, ensuring consistency across the fleet.\n\nThe administrator defines the custom metadata schema, specifying field names, types, allowed values, and whether each field is required or optional. The registry enforces these schemas for every record, guaranteeing uniformity in the information provided. Fields may be of type Text, Enum, URL, or Boolean. If a record is created without the required custom metadata, it is marked as NON_COMPLIANT and cannot be submitted for approval until all mandatory fields are filled with valid values.\n\nInvalid custom metadata, such as unknown fields, missing required fields, or invalid values, results in a ValidationException from AWS Agent Registry. When records are registered, publishers supply values for the fields specified in the chosen schema, and the registry validates these values against the administrator's schema.\n\nWith a common set of attributes, the registry aids in governance, compliance, ownership, operations, security, and compliance. Standardized information like environment (prod, staging, or dev), businessUnit, ownerTeam, supportContact, and costCenter helps determine where a resource is hosted, who is responsible for it, and who funds it. These details serve as references for configuration management databases (CMDB), cost centers, and support-tier classification.\n\nFor security and compliance purposes, data classification, compliance scopes like PCI or HIPAA, risk tiers, scan statuses, and scan report URLs are recorded. Existing security scanners, CMDB synchronization jobs, and classification tools can input these values into the registry before a record is submitted. These validated values serve as a basis for record approval and automation workflows.\n\nCustom metadata values are indexed for search, enabling consumers and agents to filter results using natural-language queries and specific metadata filters. For example, a user could retrieve approved records meeting criteria such as environment=prod AND businessUnit=payments AND securityScanStatus=PASSED. Only approved records matching the specified filters will be returned, ensuring that unpublished records with the same metadata are excluded from search results.",
  "summary": "Custom metadata schemas in AWS Agent Registry let an administrator define typed, validated fields that every agent, MCP server, and skill must carry. This post follows an MCP record from schema definition through registration, enrichment, approval, and discovery, so teams govern and find resources with one shared set of facts.",
  "key_points": [
    "AWS Agent Registry centralizes management of agents, MCP servers, and skills.",
    "Organizations can create custom metadata schemas with key-value pairs.",
    "Custom metadata enables governance, compliance, and search functionality."
  ],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}