Native Java Model Context Protocol (MCP) server for live Spring Boot inspection via Attach API.
View on GitHub ↗Native Java Model Context Protocol (MCP) server for live Spring Boot inspection. Provide Claude, Cursor, and Antigravity with real-time context on memory, threads, and database state without modifying your application code.
- The Blind AI Assistant Problem - The Solution: Zero Friction - Exposed Tools - Architectural Principles - Known Limitations and Trade-offs - Community and Contributing - License
Every backend engineer debugging complex Java applications with AI assistants hits the same wall: static source code does not reflect the live runtime state.
Existing diagnostic approaches have significant drawbacks: Python scripts analyzing manual post-mortem dumps, or Node.js tools requiring developers to modify target pom.xml files to expose Actuator endpoints or remote JMX.
JVM-MCP bridges this gap by attaching locally to the target JVM using the JDK Attach API (com.sun.tools.attach).
Distributed as a standalone native binary starting in < 15ms. No Node.js runtime required, no intermediate scripts, and critically: zero modifications to your target application code.
Once attached, the LLM gains real-time diagnostic visibility:
- Spring Boot Context: Inspect instantiated beans (listspringbeans) and identify slow bean initialization durations. - Memory & Heap: Live heap region usage, GC pause counters (getheapsummary), and top class instance counts. - Threads & Concurrency: Clean, LLM-formatted thread dumps and automated deadlock cycle detection. - HikariCP: Live connection pool metrics, connection leaks, and saturation alerts. - PostgreSQL: Direct JDBC database schema
From the project README.
Add the radar badge to your README — it shows your project was picked up by MCP Radar and links to this page:
[](https://mcp.liqiwa.com/s/oscarbol09--jvm-mcp.html)
Privacy guardrails for Spring AI: request-scoped PII tokenization and least-privilege disclosure across model/tool calls, with Presidio and JVM-local OpenNLP.
java-ai-in-action/framework-compareSpring AI vs LangChain4j vs SAA 三框架同台对比（配套篇2）
HanHan666666/patchbridge-agentLightweight browser-side AI agent framework for legacy enterprise web apps—Java 8, Spring Boot 2, TypeScript, MCP/WebMCP, OpenAI-compatible LLMs, tool calling, streaming, audit, and persistence. 面向存量企
TikyParkinson/mcp-agent-guardrails-spring-boot-starterSecurity & governance guardrails for MCP agents in Java — audit trail, agent-to-tool authorization, prompt-injection detection and rate limiting as a zero-config Spring Boot starter.
twz007/java-obs-agentFramework-agnostic Java observability sidecar agent: chat with a live JVM via MCP/web UI (metrics, slow calls, errors, logs, thread dumps, method traces). Self-hosted, attachable (Arthas-style). Apach
0xwilliamortiz/openclaude-improvedruns anywhere. uses anything
The top new MCP servers of the week, every Monday. No spam, unsubscribe anytime.