Log Analyzer

This MCP server enables AI-powered log analysis by automatically detecting common log formats, enabling smart search, error extraction, cross-file correlation, and sensitive data detection. You connect it to your Claude Code workflow, let it parse and query large log files efficiently, and get actionable debugging insights without manual parsing steps.

How to use

You use this MCP server by running it through your MCP client (such as Claude Code) and issuing natural language or command-based requests. Start by installing the MCP-enabled tool, then tell your client what log file to analyze and what you want to accomplish. You can ask it to analyze a specific log file, search for patterns with context, extract and group errors, or generate a debugging summary. The server will auto-detect formats across 9+ common log types, perform cross-file correlation, and scan for sensitive information.

Practical usage patterns include: - Analyze /var/log/app.log and tell me what’s causing the errors - Search for a term like "timeout" with context lines - Extract and group similar errors from a log file - Ask questions in natural language such as "what errors happened today?"

How to install

Prerequisites: you need Python and/or a compatible MCP client environment, plus the MCP runtime tooling (uvx) installed and available in your PATH.

Step-by-step installation and setup:

pip install codesdevs-log-analyzer

uv tool install codesdevs-log-analyzer

# Optional: configure your Claude Code settings to point to the MCP server
# Add the following to your Claude Code config file (e.g., ~/.claude/settings.json):
{
  "mcpServers": {
    "log_analyzer": {
      "command": "uvx",
      "args": ["codesdevs-log-analyzer"]
    }
  }
}

Configuration and usage notes

The MCP server supports auto-detection of multiple log formats and provides tools for parsing, searching, extracting errors, correlating events, and scanning for sensitive information. You can run it locally through a standard MCP client and point it at your log files or streams.

Examples

Find errors in a log file and group similar ones: Extract all errors from /var/log/app.log, group similar ones.

Search with context: Search for "timeout" in app.log with 5 lines of context.

Correlate events: What happened 60 seconds before each OutOfMemoryError?

Scan for secrets: Check /var/log/app.log for accidentally logged credentials.

Security and best practices

Limit access to log files and MCP endpoints to trusted users. Use secure channels when transmitting logs and avoid logging sensitive data in plain text where possible.

Troubleshooting

If the MCP client cannot connect, verify that the uvx command is available in your environment and that the CodesDevs log-analyzer tool is installed correctly. Check that your Claude Code settings.json contains the correct mcpServers entry and that the path to the log file is valid.

Notes

This MCP server supports streaming analysis for large files and aims to keep memory usage low by processing data in a streaming fashion.

Available tools

log_analyzer_parse

Detects log format and extracts metadata from the input log stream.

log_analyzer_search

Searches the log with context lines, supporting pattern matching and time filtering.

log_analyzer_extract_errors

Extracts and groups errors, capturing related stack traces.

log_analyzer_summarize

Generates a concise debugging summary from logs.

log_analyzer_correlate

Finds related events across multiple files or streams.

log_analyzer_watch

Monitors for new log entries in real time.

log_analyzer_ask

Allows natural language questions to query logs.

log_analyzer_scan_sensitive

Detects PII, credentials, and secrets in logs.

log_analyzer_another

Additional log-analysis capability to cover more scenarios.