An unofficial MCP server for using Cloud Automator REST API.
The server defines 12 tools with basic schemas and bilingual (English/Japanese) descriptions. However, critical gaps limit quality: (1) descriptions are present but lack depth and context guidance; (2) parameters have type definitions but no descriptions for what they control (e.g., 'pageSize' lacks guidance on expected range or default behavior); (3) no output schemas are documented, LLMs cannot predict response structure; (4) error handling is generic ('Couldn't list jobs') without recovery guidance; (5) all tools are READ_ONLY, which is appropriate, but no tool annotations (readOnlyHint) are declared in the schema to signal this to clients; (6) no parameter validation rules or constraints are visible (e.g., is pageSize 1 - 100? 1 - 1000?). The naming is consistent (verb_noun: list-*, get-*) and follows verb-first convention, which is a strength. However, without parameter descriptions, output schemas, and contextual help, LLMs must guess parameter intent and cannot plan downstream operations confidently.
Get an AWS account which belongs to the group with a specific ID. グループに属する指定されたIDのAWSアカウントを取得。
Get a Google Cloud account which belongs to the group with a specific ID. グループに属する指定されたIDのGoogle Cloudアカウントを取得。
Get a job with a specific ID. 指定されたIDのジョブを取得。
Get a job workflow with a specific ID. 指定されたIDのジョブワークフローを取得。
Get a job log with a specific ID. 指定されたIDのジョブログを取得。
Get a post process with a specific ID. 指定されたIDの後処理を取得。
No parameter descriptions. Parameters like 'pageNumber', 'pageSize', 'jobId', 'logId' have type but no description explaining what they control, expected range, or defaults. LLMs cannot infer meaning from names alone.
No output schemas documented. Tool descriptions do not specify what fields are returned, their types, or structure. LLMs cannot predict response shape or plan downstream tool calls confidently.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-22 | D | 50 | 2026-07-28+ | v2 |
| 2026-03-09 | F | 45 | - | v1 |
Get resource operation results which belongs to the job log with a specific ID. 指定されたIDのジョブログのリソース操作結果を取得。
List all AWS accounts with pagination. グループに属するAWSアカウントの一覧を取得(ページネーションあり)。
List all Google Cloud accounts with pagination. グループに属するGoogle Cloudアカウントの一覧を取得(ページネーションあり)。
List all job workflows with pagination. ジョブワークフローの一覧を取得(ページネーションあり)。
List all jobs with pagination. ジョブの一覧を取得(ページネーションあり)。
List all post processes with pagination. 後処理の一覧を取得(ページネーションあり)。
Generic error messages without recovery guidance. All tools return 'Couldn't list [resource]' or 'Couldn't get [resource]' with no indication of what to try next, whether the error is retryable, or how to validate input.
No parameter constraints documented. Pagination parameters (pageNumber, pageSize) lack stated ranges (min, max, defaults). Are they 1-indexed or 0-indexed? Is pageSize bounded? Without constraints, LLMs may pass invalid values.
No tool annotations for read-only intent. All 12 tools are marked Risk: READ_ONLY in metadata, but readOnlyHint is not declared in the tool definition. Clients cannot distinguish read-only from destructive tools.
Descriptions are bilingual but lack actionable context. While 'List all jobs with pagination' is clear, there is no guidance on when to call list-jobs vs get-job, what structure is returned, or how to use pagination to iterate large datasets.