An MCP server for monitoring and diagnosing Spring application servers. Provides tools to check server health status, analyze error logs, fetch recent logs, register new servers, and get setup guides.
This server has mixed quality. Tool names use clear verb-noun patterns (list_, diagnose_, fetch_, get_, register_) and parameters have type definitions. However, descriptions are bilingual (Korean + English) which inflates character count without clarity, some parameter descriptions lack actionable detail, output schemas are completely undocumented, and error handling guidance is absent. All 7 tools ARE explicitly registered with @McpTool annotations in ServerDoctorMcpTools.java, so inference penalty does not apply. The schema quality is moderate, basic types are present but constraints, ranges, and response structures are nowhere to be found.
Spring 서버 진단 MCP가 대상 서버의 최근 에러 로그와 리소스 상태를 조회하여 종합적으로 분석합니다. (EN) Spring Server Diagnostics MCP comprehensively analyzes recent error logs and resource status of the target server.
Spring 서버 진단 MCP가 서버에서 최근 발생한 에러 로그들을 조회 후 분석합니다. (EN) Spring Server Diagnostics MCP fetches and analyzes recent error logs from the server.
Spring 서버 진단 MCP가 서버의 최근 로그를 INFO/WARN/ERROR 레벨 구분 없이 최신순으로 조회합니다. 에러 외 흐름을 파악하거나 장애 전후 맥락을 확인할 때 사용하세요. (EN) Spring Server Diagnostics MCP fetches the server's recent logs in latest-first order across INFO/WARN/ERROR levels. Use it to understand flow beyond errors or the context around an incident.
Spring 서버 진단 MCP가 서버의 현재 Health 상태(UP/DOWN), 마지막 체크 시각, latency, stale 여부(60초 기준)를 반환합니다. (EN) Spring Server Diagnostics MCP returns the server's current health status (UP/DOWN), last check time, latency, and staleness (60s threshold).
Spring 서버 진단 MCP 모니터링을 시작하는 방법을 보여줍니다. Application.yml, docker-compose.yml, .env 설정에 대해 알려줍니다. (EN) Shows how to start monitoring with Spring Server Diagnostics MCP, covering application.yml, docker-compose.yml, and .env configuration.
No output schemas documented. Tools return String responses (e.g. 'String diagnoseServer(...)') but there is no specification of the JSON structure, field names, or data types the LLM should expect. This forces LLMs to parse unstructured text, increasing hallucination and token waste.
Parameter descriptions lack actionable constraints and ranges. 'sinceHours' accepts Integer but has no min/max bounds stated in the description (only a vague '기본값 24'). LLMs cannot infer whether 1000 or -5 are valid. The description should say 'Integer, minimum 1, maximum 365, default 24.'
Bilingual descriptions inflate character count without improving clarity for LLMs. Korean + English in one description wastes ~50% of the token budget. Use a single language (preferably English for LLM optimization) and provide localization outside the tool definition.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-22 | F | 43 | 2026-07-28+ | v2 |
| 2026-03-09 | F | 45 | - | v1 |
Spring 서버 진단 MCP에서 PlayMCP 원격 환경에서 즉시 체험 가능한 데모 서버 목록을 반환합니다. 실제 사용자 서버 목록은 반환하지 않습니다. (EN) Returns the list of demo servers available for instant trial on Spring Server Diagnostics MCP via PlayMCP. Does not return real user servers.
Spring 서버 진단 MCP에 모니터링할 새로운 대상 서버를 등록하고, 연동 가이드(yml, env 등)를 생성합니다. (EN) Registers a new target server to monitor on Spring Server Diagnostics MCP and generates an integration guide (yml, env, etc).
No error recovery guidance. If a tool call fails (e.g. server not found, connection timeout), the tool returns a String error message with no indication of whether the error is retryable, which tool to call next, or what the LLM should ask the user. See pattern:recovery-guide.
No pagination or result limits documented. Tools like fetch_error_logs and fetch_recent_logs could return hundreds of log entries, but there is no description of max results, pagination parameters, or how LLMs should handle large result sets.