Journald MCP server
Incident forensic with log files analyzing
Journald MCP Server
An MCP server for accessing systemd journal logs.
Features
- List systemd units from journal logs
- List syslog identifiers from journal logs
- Get datetime of first journal entry
- Filter journal entries by datetime range (since/until)
- Filter by systemd unit or syslog identifier
- Filter by message content (case-insensitive substring matching)
- Natural language datetime parsing (e.g., "2 hours ago", "yesterday at 3pm")
- List units and identifiers within specific time ranges
Installation
# Install dependencies
uv sync
Usage
Run as non-root: Give the user systemd-journal group access usermod -aG systemd-journal $USER
Run the server with:
uv run server.py [OPTIONS]
CLI Options
--transport: Transport protocol to use (stdio,sse, orstreamable-http). Default:stdio--port: Port to listen on for HTTP transport (ignored forstdiotransport). Default:3002--log-level: Logging level (DEBUG,INFO,WARNING,ERROR,CRITICAL). Default:INFO
Examples
-
Run with stdio transport (default, for MCP clients that communicate via stdin/stdout):
python server.py -
Run with HTTP transport on custom port:
python server.py --transport streamable-http --port 8080 -
Run with SSE transport:
python server.py --transport sse --port 3000 -
Run with debug logging:
python server.py --log-level DEBUG
MCP Integration
The server provides the following MCP resources and tools:
Resources
journal://units: List unique systemd units from journal logs (all accessible time)journal://syslog-identifiers: List unique syslog identifiers from journal logs (all accessible time)journal://first-entry-datetime: Get the datetime of the first entry in the journaljournal://units/{since}/{until}: List unique systemd units within a specified time rangejournal://syslog-identifiers/{since}/{until}: List unique syslog identifiers within a specified time range
Tools
-
get_journal_entries: Get journal entries with datetime filtering- Parameters:
since(optional),until(optional),unit(optional),identifier(optional),message_contains(optional),limit(default: 100) - Returns: List of entries with timestamp, unit, identifier, and message
- Example: Get logs from last 2 hours containing "error":
since="2 hours ago", message_contains="error"
- Parameters:
-
get_recent_logs: Get recent journal logs from the last N minutes- Parameters:
minutes(default: 60),unit(optional),limit(default: 50) - Returns: Formatted string of recent log messages
- Parameters:
Datetime Input Format
The server uses natural language datetime parsing via the dateparser library. Supported formats include:
- Relative times: "2 hours ago", "yesterday at 3pm", "last week", "now"
- Absolute times: "2024-01-15 14:30", "2024-01-15T14:30:00"
- Mixed: "today at 9am", "tomorrow 3pm"
All times are interpreted as UTC and returned in human-readable format: "YYYY-MM-DD HH:MM:SS UTC"
Development
This project uses:
- Python 3.12+
- MCP FastMCP
- systemd-python for journal access
- Click for CLI interface
- dateparser for natural language datetime parsing
Project Structure
journald-mcp-server/
├── journald_mcp_server/ # Main package
│ ├── __init__.py
│ ├── server.py # MCP server implementation
│ └── datetime_utils.py # Datetime parsing and formatting utilities
├── tests/ # Test suite
│ ├── __init__.py
│ └── test_server.py
├── server.py # Entry point wrapper
├── pyproject.toml
└── README.md
Running Tests
python -m pytest tests/
Serveurs connexes
Market Intelligence MCP
Recession probability scoring, capital rotation analysis, investment calculators, and real-time economic data. 19 tools for financial market intelligence
PixelPanda
31 AI image processing tools — free local editing, AI background removal & upscaling, and paid product photo generation
Text-to-Speech (TTS)
A Text-to-Speech server supporting multiple backends like macOS say, ElevenLabs, Google Gemini, and OpenAI TTS.
Discode (Code Mode MCP)
Code Mode enabled MCP server - save 90% of token usage for any MCP-compatible agent
Firelinks.cc MCP
Create and manage short links for tracking and distributing traffic.
Universal Image MCP
Universal MCP server for AI image generation supporting AWS Bedrock (Nova Canvas), OpenAI (GPT Image, DALL-E), and Google Gemini (Imagen 4). Generate, transform, and edit images using multiple AI models through a single Model Context Protocol interface.
MCP Location Services for LLM
Provides geocoding, reverse geocoding, routing, and driving distance calculations for LLMs using pgGeocoder and pgrServer.
Trayd
Trade Robinhood with natural language in Claude Code
Nomad Stays
The world's platform for finding and booking digital nomad friend accommodation
LLM Brand Monitor MCP Server
MCP server for LLM Brand Monitor — track how AI models mention your brand. Requires an API key (LBM_API_KEY) from llmbrandmonitor.com.