JVM Flight Recorder (JFR) Forensic Engine for LLM
Hybrid TypeScript/Java MCP Server for Advanced JVM Analysis
Lumen MCP is the analytical brain of the Lumen ecosystem. It leverages a high-performance Java engine to parse JFR files and a TypeScript bridge to provide a seamless Model Context Protocol (MCP) interface for AI agents like Gemini and Claude.
- Hybrid Intelligence: Combines Java's deep JFR parsing power with Node.js's ecosystem compatibility.
- Exception Forensics: Trace hidden exceptions (Swallowed Exceptions) with full call chains.
- SQL & I/O Analytics: Pinpoint slow queries and socket latencies with line-level precision.
- Smithery Ready: Fully compatible with Smithery CLI for instant distribution.
Lumen MCP requires both Node.js (18+) and Java (11+) to build.
npm install
npm run build(This command runs tsc for the bridge and ./gradlew installDist for the engine.)
You can now run Lumen using the Node.js bridge (Recommended):
{
"mcpServers": {
"lumen": {
"command": "node",
"args": [
"/path/to/lumen-mcp/dist/server.js"
]
}
}
}| Tool | Description |
|---|---|
analyze_exceptions |
Analyzes exception throw points and call chains. |
analyze_jdbc_queries |
Identifies slow SQL queries and their source lines (Java 11+). |
analyze_network_io |
Detects socket latency (Essential for Java 8 triage). |
analyze_hot_methods |
Pinpoints code lines consuming the most CPU cycles. |
analyze_memory_usage |
Analyzes GC health and object allocation hotspots. |
analyze_lock_contention |
Identifies thread synchronization bottlenecks. |
"Gemini, use lumen to analyze
data/incident.jfr. Tell me why the database response is slow."
© 2026 SnipeFactory. All rights reserved.