Our catalog surfaces 199 AWS-related matches. This guide separates product association from publisher identity, then connects each implementation to distribution and current security evidence.
The catalog’s 199 AWS matches include implementations that connect to AWS products, AWS-associated packages, and community projects. Brand association alone does not establish publisher identity, security review, or endorsement.
Search results frequently collapse these populations together. They have different ownership, hosting, update, authentication, and evidence models.
| Population | Who operates it | Typical delivery | What to verify |
|---|---|---|---|
| AWS-managed MCP Server | AWS-managed service | Remote regional endpoint | IAM and SCP scope, proxy configuration, mutating access, CloudTrail coverage |
| AWS Labs open-source servers | AWS Labs repositories and contributors | Primarily locally run packages; varies by server | Exact repository and package, credentials used, tool authority, maintenance state |
| Community AWS servers | Independent publishers | npm, PyPI, source, or remote endpoint | Publisher provenance, package integrity, remote operator, permissions, verification evidence |
The decisive variable is not the AWS label. It is the combination of operator identity, execution boundary, and effective authority.
Verify the repository owner, package publisher, remote domain, and whether first-party status is supported by retained evidence.
Map every exposed tool to the IAM principal, resources, regions, and mutating operations available in the deployment.
Use exact-version verification as bounded evidence. Missing evidence means unknown; a clean retained run is not certification.
Counts below are catalog search appearances and can overlap. Open a result set to evaluate individual implementations rather than treating the total as a list of official servers.
| Product family | Catalog matches | Integration surface | Complete result set |
|---|---|---|---|
| AWS | 199 | Cloud infrastructure, deployment, operations, data, and AI services | View all AWS results → |
| Amazon | 81 | Cross-product records explicitly referencing Amazon | View all Amazon results → |
| Amazon Bedrock | 4 | Foundation models, agents, knowledge bases, and generative AI | View all Amazon Bedrock results → |
| Amazon Redshift | 12 | Cloud data warehouse querying, analytics, and administration | View all Amazon Redshift results → |
| Amazon CloudWatch | 8 | Metrics, logs, alarms, and operational observability | View all Amazon CloudWatch results → |
| Amazon DynamoDB | 6 | NoSQL tables, records, queries, and cloud data operations | View all Amazon DynamoDB results → |
These records demonstrate why attribution and verification must remain separate. They are examples, not a complete or ranked recommendation list.
| Implementation | Distribution | Current security evidence | Evidence record |
|---|---|---|---|
| Amazon All-in-One Scrape MCPAmazon | Distribution unresolved | Verification pendingNo completed public current-version result is available. | Open evidence → |
| Amazon DataZone MCP ServerAmazon DataZone | PyPI · amazon-datazone-mcp-server · 0.1.1 | Verification pendingNo completed public current-version result is available. | Open evidence → |
| AWS Nova CanvasAWS Nova Canvas | PyPI · 1.1.0 | Verification completedNo proven finding in the retained current-version test scope. | Open evidence → |
| Yawlabs AWS MCPAWS; Yawlabs named publisher | npm · @yawlabs/aws-mcp · 1.8.2 | Verification completedNo proven finding in the retained current-version test scope. | Open evidence → |
AWS describes its managed MCP Server as a remote service that gives compatible agents authenticated AWS access through a small tool set. It is distinct from the collection of open-source AWS Labs servers and from third-party servers that mention AWS.
No. An AWS product name can describe the service an MCP connects to without proving that AWS published, reviewed, or endorsed the implementation. Publisher identity must be verified separately.
No server is safe by name alone. Review publisher evidence, authentication, effective IAM permissions, mutating tools, remote endpoints, auditability, package provenance, and version-specific verification before adoption.
AWS Labs lists dozens of specialized AWS servers, while this catalog currently surfaces 199 AWS search matches. The catalog count is a discovery measure, not a deduplicated count of official AWS servers.
Method. Catalog counts were captured on 31 August 2026. The 310 appearances across all product rows overlap and are not a deduplicated implementation count; “199 AWS matches” is the direct AWS query count. Association is based on product terms in catalog identity and description fields and does not imply first-party publication, endorsement, or certification. Verification applies only to the selected exact version and retained test scope. Search every Amazon and AWS catalog match →