tda

Overview

TDA (Thread Dump Analyzer) is a comprehensive Java application that provides both a Swing GUI and an MCP server for analyzing Java thread dumps. It can detect deadlocks, identify long-running threads, and analyze virtual threads with carrier thread pinning. As an MCP server, it enables AI tools like Cursor and Claude Desktop to perform headless analysis of thread dumps without requiring manual inspection of raw stack traces.

Try asking AI

After installing, here are 5 things you can ask your AI assistant:

• parse_logpath (string, required)

  Parses a log file containing Java thread dumps.

  Try:Parse this Java thread dump log fileTry:Analyze the thread dump log at /path/to/thread.log
• get_summary

  Returns a summary of all parsed thread dumps.

  Try:Get a summary of all thread dumpsTry:Show me the overview of the thread analysis
• check_deadlocks

  Checks for and returns information about deadlocks in parsed thread dumps.

  Try:Check for deadlocks in the thread dumpsTry:Are there any deadlocks detected?
• find_long_running

  Identifies threads that remain in the same state across consecutive dumps.

  Try:Find threads that have been running too longTry:Which threads are stuck in the same state?
• analyze_virtual_threads

  Detects virtual threads where the carrier thread is stuck.

  Try:Analyze virtual thread issuesTry:Find problems with virtual threads and carrier threads
• get_native_threadsdump_index (int, required)

  Returns threads in native methods for a specific thread dump.

  Try:Get native threads in dump 3Try:Show threads using native methods
• get_zombie_threads

  Returns zombie threads (unresolved SMR addresses) with timestamps.

  Try:Find zombie threads in the dumpsTry:Are there any zombie threads?
• clear

  Resets the server state and clears the thread store.

  Try:Clear the current analysis and start freshTry:Reset the thread analyzer

Comparable tools

Installation

java -Djava.awt.headless=true -jar tda.jar --mcp

{
  "mcpServers": {
    "tda": {
      "command": "java",
      "args": ["-Djava.awt.headless=true", "-jar", "/path/to/tda.jar", "--mcp"]
    }
  }
}

FAQ

What types of thread dumps does TDA support?

TDA supports traditional text-based thread dumps and JSON thread dumps (experimental). It works with Java 1.4.x through Java 21+, including specialized support for Virtual Threads (Java 19+).

How does TDA integrate with AI tools?

TDA runs as an MCP server in headless mode, exposing tools like parse_log, check_deadlocks, find_long_running, and analyze_virtual_threads. AI tools can use these to analyze thread dumps without directly reading large log files.

Compare tda with