Loading ToolBench
MCP server for accelerating Docker image pulls from overseas registries using GitHub Actions and hub-mirror
Three tools with complete schemas and descriptions, but significant quality gaps. Tool names lack action verbs (verify_engine_status, submit_image_pull_task, get_task_status are acceptable but generic). Descriptions are present (avg ~100 chars) but lack WHEN/WHY guidance and dependency hints. Parameters have types and descriptions, but no enums for constrained values. No output schemas documented. Error handling returns plain text without recovery guidance. Chinese descriptions reduce LLM clarity. No tool annotations (readOnlyHint/destructiveHint). Missing idempotency guarantees and pagination for list-like operations.
根据任务 ID 查询进度。如果提供 original_images 参数，成功时将返回一键拉取和重命名的完整命令。
异步提交镜像加速拉取任务。此工具会立即返回任务 ID，不会阻塞等待。
检查镜像加速引擎的配置和连接状态是否正常。建议在拉取镜像前先进行此检查。
No output schemas documented. LLMs cannot infer what fields to expect from tool responses, forcing unstructured parsing and breaking downstream tool chaining.
Descriptions in Chinese reduce clarity for English-speaking LLMs and violate the pattern of writing descriptions as prompt-engineering guidance. Descriptions lack WHEN/WHY context and dependency hints.
No tool annotations (readOnlyHint, destructiveHint, idempotentHint). submit_image_pull_task is marked WRITE but lacks destructiveHint annotation. Agents cannot determine retry safety.
Error responses return plain text without recovery guidance. 'Unknown tool' and generic error messages do not tell LLMs what to do next or whether to retry.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | C | 62 | 2026-07-28+ | v2 |
platform parameter in submit_image_pull_task accepts free-form strings with no enum constraint. Should define valid architectures (linux/amd64, linux/arm64, etc.) to prevent hallucinated values.