No user-submitted related servers found.
Scored across 69 tools
Several tool pairs overlap significantly: build_server_memory and refresh_server_memory are explicitly functionally equivalent, ip_ban_set and block_ip both ban IPs via WAF/SG/NACL, and check_status vs get_server_info vs get_server_memory all describe an instance at different granularities. The verbose cross-referencing descriptions save it from a lower score, but an agent could still easily misselect between these pairs.
Snake_case verb_noun is the dominant pattern with clear prefix families (aws_, s3_, docker_, cloudwatch_, db_), making the namespace highly predictable and scannable. Minor deviations: build_server_memory vs refresh_server_memory are functionally identical despite implying different purposes, and docker_ps breaks verb_noun in favor of the familiar CLI command.
69 tools is far too many for a single MCP surface and will strain an agent's ability to select among them reliably. The count is partially justified by the genuinely broad domain (server management, AWS, S3, Docker, DB, security), but the server would be much more coherent split into several focused servers.
The surface covers the full operational lifecycle: instance management, a complete memory build/refresh/recall system, S3 CRUD, EC2 lifecycle with all prerequisite lookups, a Docker diagnostic suite, DB setup and querying, and a coherent abuse-response workflow (enrich, describe ingress, block, rate-limit). Minor gaps exist (no non-AWS instance provisioning, no S3 policy/versioning, no general CloudWatch metrics), but aws_call and api_request act as escape hatches.