Building an MCP Server for Java Performance Profiling

Alexey Kurilov — Fri, 13 Mar 2026 06:11:17 +0000

I built javaperf - an MCP (Model Context Protocol) server that lets AI assistants profile Java applications without touching the command line.

If you've ever debugged Java performance issues, you know the drill: jps, jcmd, jfr... memorizing flags, parsing output. What if your AI assistant could do this for you?

What is MCP?

Model Context Protocol (MCP) is a standard for connecting AI assistants to external tools. Instead of AI just generating code suggestions, it can actually:

Run diagnostics on your running Java apps
Analyze thread dumps
Inspect JFR recordings
All through natural conversation

javaperf Features

Process Discovery: Auto-detect running Java processes via jps
Thread Analysis: Dump and analyze threads (jcmd Thread.print)
VM Flags: Inspect JVM configuration
JFR Support: Record and analyze Java Flight Recorder data
Performance Counters: Access low-level JVM metrics

Installation

npm install -g javaperf

Or use directly via npx (no install needed):

npx javaperf

Connecting to Claude Desktop

Add to your claude_desktop_config.json:

{
  "mcpServers": {
    "javaperf": {
      "command": "npx",
      "args": ["-y", "javaperf"]
    }
  }
}

Example Usage

Once connected, you can ask Claude things like:

"Show me all running Java processes"
"Analyze thread dump for process 1234"
"Start a 60-second JFR recording"
"Why is my app using so much memory?"

The AI handles the jcmd, jfr, and jps commands behind the scenes.

Why This Matters

Before MCP: You Google the right flags, copy-paste commands, manually parse output.

After MCP: "Hey Claude, why is my Spring Boot app slow?" → AI investigates → Finds thread contention → Suggests fixes.

Try It Out

If you're working with Java and using Claude Desktop or Cursor IDE, give it a spin. Would love feedback on what other profiling features would be useful!

Built with TypeScript, powered by JDK tools you already have.

I Built a Self-Hosted Web UI for 11 Databases — Here's Why You Might Want One Too

Alexey Kurilov — Mon, 09 Mar 2026 15:47:37 +0000

👋 The Problem

If you work with microservices in Kubernetes, you know the pain:

Your app uses PostgreSQL for transactions, Redis for cache, Kafka for events, and MongoDB for documents.
To debug a production issue, you open: psql, redis-cli, kcat, mongosh, maybe curl for Elasticsearch...
You copy-paste IDs between terminals. You forget which environment you're in. You accidentally run DROP on staging.

I've been there. So I built Panopticum — a lightweight, self-hosted web UI that lets you browse, query, and edit data across 11+ database types from a single interface. No Electron, no heavy desktop app, no CDN dependencies.

🛠️ Why Another Database Tool?

Existing tools are great, but they didn't fit my workflow:

Tool	Gap I Felt
DBeaver / DataGrip	Desktop-only, heavy, per-connection setup
Adminer / phpMyAdmin	Web-based, but single-DB-type, PHP stack
pgAdmin / Mongo Express	Excellent for one DB, but I needed many
Custom internal tools	Took weeks to build; not reusable

I wanted something that could be:
✅ Deployed in one command to any K8s cluster

✅ Secure by default (HTTP Basic + env vars)

✅ Fully offline / air-gapped (no external CDNs)

✅ Extendable via REST API

✅ Simple enough for QA engineers, powerful enough for devs

So I built it with Micronaut 4, HTMX, and Thymeleaf — and open-sourced it.

🚀 Quick Start

# Run locally
./gradlew run

# Or with Docker
docker run -d --name panopticum \
  -p 8080:8080 \
  -v panopticum-data:/data \
  -e PANOPTICUM_USER=admin \
  -e PANOPTICUM_PASSWORD=changeme \
  ghcr.io/thesharque/panopticum:latest

Open http://localhost:8080 — and you're in.

🔌 Supported Databases (All in One Place)

Type	Browse	Query	Edit Rows
PostgreSQL / CockroachDB / YugabyteDB	✅	✅ SQL	✅ (by ctid)
MySQL / MariaDB	✅	✅ SQL	✅ (with PK/unique)
MS SQL Server	✅	✅ SQL	✅ (with PK/unique)
Oracle	✅	✅ SQL	✅ (by ROWID)
MongoDB	✅	✅ Query JSON	✅
Redis / Dragonfly / Valkey	✅	—	✅ (key/value)
ClickHouse	✅	✅ SQL	—
Cassandra / ScyllaDB	✅	✅ CQL	✅ (with PK)
RabbitMQ	✅ queues	—	❌ read-only
Kafka	✅ topics	—	❌ read-only
Elasticsearch / OpenSearch	✅ indices	✅ Query DSL	✅ (by `_id`)

💡 All connections are stored in an embedded H2 database. Add/remove them via UI or bootstrap via PANOPTICUM_CONNECTIONS_JSON env var.

✨ Features That Made My Life Easier

🔐 Zero-Config Auth & Offline-First

HTTP Basic Auth (credentials from env vars — easy with K8s Secrets)
All frontend assets (HTMX, CodeMirror, Prism) bundled locally — no CDN, works in air-gapped clusters
Light/dark theme, EN/RU localization

⚡ HTMX for Snappy UX

No React, no build step for the frontend. Partial page updates via HTMX keep the UI responsive while keeping the codebase simple and maintainable.

<!-- Example: Load table rows without full reload -->
<div hx-get="/api/postgresql/connection-1/schema/public/table/users"
     hx-trigger="load"
     hx-target="#rows-container">
  <div id="rows-container">Loading...</div>
</div>

🔄 Data Diff: Compare Records Across Environments

This is my favorite feature. Found a suspicious record in Prod?

Click "Add to Compare" on any row detail page
Switch to Stage or Dev connection
Open the same logical record
Click "Compare" — see side-by-side JSON/field diff

All stored in your browser's localStorage — no server roundtrip. Perfect for debugging environment drift.

🌐 Full REST API + Swagger

Every UI action has a corresponding API endpoint. Documented interactively at /swagger-ui:

# Execute SQL via API
curl -u admin:changeme http://localhost:8080/api/postgresql/conn-1/query \
  -H "Content-Type: application/json" \
  -d '{"sql": "SELECT id, email FROM users WHERE status = $1", "params": ["active"]}'

Great for automation, CI checks, or building custom dashboards.

🧭 Who Is This For?

DevOps/SREs: Debug data issues in staging/prod without installing 5 CLI tools
QA Engineers: Validate test data across services with a simple UI
Backend Developers: Quick ad-hoc queries without leaving the browser
Security Teams: Audit access via basic auth + logs; no external dependencies

👉 Check it out, star, or open an issue:

https://github.com/theSharque/panopticum

💬 Let's Chat

What's your biggest pain point when debugging across multiple data stores?
Would you use a tool like this in your workflow?
Any databases you'd like to see added?

Drop a comment below — I read every one. 🙏

DEV Community: Alexey Kurilov