MCP 与技能管理面板:设置页展示 MCP 服务器与 Skill 目录,随时启停释放上下文占用;可选 AI 中间层(dsh_mcp_search/dsh_mcp_call)按需使用 MCP。
Both tools have minimal descriptions (Chinese text, ~30-40 chars each) and lack visible input schema definitions in the provided source. Tool names use underscores but lack clear action verbs (dsh_mcp_search, dsh_mcp_call are meta-tools, not domain-specific). No parameter type information visible. No output schema documented. The tools appear to be middleware/gateway tools rather than domain-specific agents, which limits their utility for LLM planning. Critical: cannot verify schema presence in source code provided.
MCP 工具调用(中间层控制工具)
MCP 工具搜索与能力检索(中间层控制工具)
No input schemas visible in source code. Tools declare parameters (query, limit, toolName, arguments) but no JSON Schema type definitions, constraints, or validation rules are shown.
Descriptions are under 50 characters and lack actionable context. 'MCP 工具搜索与能力检索(中间层控制工具)' does not explain WHEN to use this tool, what it returns, or how it differs from direct MCP calls.
No output schema documented. LLMs cannot plan downstream calls without knowing what fields dsh_mcp_search returns or what structure dsh_mcp_call produces.
Tool names lack clear action verbs. 'dsh_mcp_search' and 'dsh_mcp_call' are meta-operations on MCP itself, not domain-specific actions. Names do not convey intent to LLMs (e.g., 'search_mcp_tools', 'invoke_mcp_tool' would be clearer).
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 27 | 2026-07-28+ | v2 |
No error handling guidance. If dsh_mcp_call fails (tool not found, invalid arguments), no recovery path is documented for the LLM.