Loki MCP Server
An MCP ( Model Context Protocol ) Server for Grafana Loki
scottlepp
README
Loki MCP Server
A Go-based server implementation for the Model Context Protocol (MCP) with Grafana Loki integration.
Getting Started
Prerequisites
- Go 1.16 or higher
Building and Running
Build and run the server:
# Build the server
go build -o loki-mcp-server ./cmd/server
# Run the server
./loki-mcp-server
Or run directly with Go:
go run ./cmd/server
The server communicates using stdin/stdout following the Model Context Protocol (MCP). This makes it suitable for use with Claude Desktop and other MCP-compatible clients. It does not run as an HTTP server on a port.
Project Structure
.
├── cmd/
│ ├── server/ # MCP server implementation
│ └── client/ # Client for testing the MCP server
├── internal/
│ ├── handlers/ # Tool handlers
│ └── models/ # Data models
├── pkg/
│ └── utils/ # Utility functions and shared code
└── go.mod # Go module definition
MCP Server
The Loki MCP Server implements the Model Context Protocol (MCP) and provides the following tools:
Loki Query Tool
The loki_query
tool allows you to query Grafana Loki log data:
-
Required parameters:
query
: LogQL query string
-
Optional parameters:
url
: The Loki server URL (default: from LOKI_URL environment variable or http://localhost:3100)start
: Start time for the query (default: 1h ago)end
: End time for the query (default: now)limit
: Maximum number of entries to return (default: 100)
Environment Variables
The Loki query tool supports the following environment variables:
LOKI_URL
: Default Loki server URL to use if not specified in the request
Testing the MCP Server
You can test the MCP server using the provided client:
# Build the client
go build -o loki-mcp-client ./cmd/client
# Loki query examples:
./loki-mcp-client loki_query "{job=\"varlogs\"}"
./loki-mcp-client loki_query "http://localhost:3100" "{job=\"varlogs\"}"
./loki-mcp-client loki_query "{job=\"varlogs\"}" "-1h" "now" 100
# Using environment variable:
export LOKI_URL="http://localhost:3100"
./loki-mcp-client loki_query "{job=\"varlogs\"}"
Docker Support
You can build and run the MCP server using Docker:
# Build the Docker image
docker build -t loki-mcp-server .
# Run the server
docker run --rm -i loki-mcp-server
Alternatively, you can use Docker Compose:
# Build and run with Docker Compose
docker-compose up --build
Local Testing with Loki
The project includes a complete Docker Compose setup to test Loki queries locally:
-
Start the Docker Compose environment:
docker-compose up -d
This will start:
- A Loki server on port 3100
- A Grafana instance on port 3000 (pre-configured with Loki as a data source)
- A log generator container that sends sample logs to Loki
- The Loki MCP server
-
Use the provided test script to query logs:
# Make it executable chmod +x test-loki-query.sh # Run with default parameters (queries last 15 minutes of logs) ./test-loki-query.sh # Query for error logs ./test-loki-query.sh '{job="varlogs"} |= "ERROR"' # Specify a custom time range and limit ./test-loki-query.sh '{job="varlogs"}' '-1h' 'now' 50
-
Insert dummy logs for testing:
# Make it executable chmod +x insert-loki-logs.sh # Insert 10 dummy logs with default settings ./insert-loki-logs.sh # Insert 20 logs with custom job and app name ./insert-loki-logs.sh --num 20 --job "custom-job" --app "my-app" # Insert logs with custom environment and interval ./insert-loki-logs.sh --env "production" --interval 0.5 # Show help message ./insert-loki-logs.sh --help
-
Access the Grafana UI at http://localhost:3000 to explore logs visually.
Architecture
The Loki MCP Server uses a modular architecture:
- Server: The main MCP server implementation in
cmd/server/main.go
- Client: A test client in
cmd/client/main.go
for interacting with the MCP server - Handlers: Individual tool handlers in
internal/handlers/
loki.go
: Grafana Loki query functionality
Using with Claude Desktop
You can use this MCP server with Claude Desktop to add Loki query tools. Follow these steps:
Option 1: Using the Compiled Binary
- Build the server:
go build -o loki-mcp-server ./cmd/server
- Add the configuration to your Claude Desktop configuration file using
claude_desktop_config_binary.json
.
Option 2: Using Go Run with a Shell Script
- Make the script executable:
chmod +x run-mcp-server.sh
- Add the configuration to your Claude Desktop configuration file using
claude_desktop_config_script.json
.
Option 3: Using Docker (Recommended)
- Build the Docker image:
docker build -t loki-mcp-server .
- Add the configuration to your Claude Desktop configuration file using
claude_desktop_config_docker.json
.
Configuration Details
The Claude Desktop configuration file is located at:
- On macOS:
~/Library/Application Support/Claude/claude_desktop_config.json
- On Windows:
%APPDATA%\Claude\claude_desktop_config.json
- On Linux:
~/.config/Claude/claude_desktop_config.json
You can use one of the example configurations provided in this repository:
claude_desktop_config.json
: Generic templateclaude_desktop_config_example.json
: Example usinggo run
with the current pathclaude_desktop_config_binary.json
: Example using the compiled binaryclaude_desktop_config_script.json
: Example using a shell script (recommended forgo run
)claude_desktop_config_docker.json
: Example using Docker (most reliable)
Notes:
-
When using
go run
with Claude Desktop, you may need to set several environment variables in both the script and the configuration file:HOME
: The user's home directoryGOPATH
: The Go workspace directoryGOMODCACHE
: The Go module cache directoryGOCACHE
: The Go build cache directory
These are required to ensure Go can find its modules and build cache when run from Claude Desktop.
-
Using Docker is the most reliable approach as it packages all dependencies and environment variables in a container.
Or create your own configuration:
{
"mcpServers": {
"lokiserver": {
"command": "path/to/loki-mcp-server",
"args": [],
"env": {
"LOKI_URL": "http://localhost:3100"
},
"disabled": false,
"autoApprove": ["loki_query"]
}
}
}
Make sure to replace path/to/loki-mcp-server
with the absolute path to the built binary or source code.
-
Restart Claude Desktop.
-
You can now use the tools in Claude:
- Loki query examples:
- "Query Loki for logs with the query {job="varlogs"}"
- "Find error logs from the last hour in Loki using query {job="varlogs"} |= "ERROR""
- "Show me the most recent 50 logs from Loki with job=varlogs"
- Loki query examples:
Using with Cursor
You can also integrate the Loki MCP server with the Cursor editor. To do this, add the following configuration to your Cursor settings:
Docker configuration:
{
"mcpServers": {
"loki-mcp-server": {
"command": "docker",
"args": ["run", "--rm", "-i", "-e", "LOKI_URL=http://host.docker.internal:3100", "loki-mcp-server:latest"]
}
}
}
After adding this configuration, restart Cursor, and you'll be able to use the Loki query tool directly within the editor.
License
This project is licensed under the MIT License - see the LICENSE file for details.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
MCP Package Docs Server
Facilitates LLMs to efficiently access and fetch structured documentation for packages in Go, Python, and NPM, enhancing software development with multi-language support and performance optimization.
Claude Code MCP
An implementation of Claude Code as a Model Context Protocol server that enables using Claude's software engineering capabilities (code generation, editing, reviewing, and file operations) through the standardized MCP interface.
@kazuph/mcp-taskmanager
Model Context Protocol server for Task Management. This allows Claude Desktop (or any MCP client) to manage and execute tasks in a queue-based system.
Linear MCP Server
Enables interaction with Linear's API for managing issues, teams, and projects programmatically through the Model Context Protocol.
mermaid-mcp-server
A Model Context Protocol (MCP) server that converts Mermaid diagrams to PNG images.
Jira-Context-MCP
MCP server to provide Jira Tickets information to AI coding agents like Cursor

Linear MCP Server
A Model Context Protocol server that integrates with Linear's issue tracking system, allowing LLMs to create, update, search, and comment on Linear issues through natural language interactions.

Sequential Thinking MCP Server
This server facilitates structured problem-solving by breaking down complex issues into sequential steps, supporting revisions, and enabling multiple solution paths through full MCP integration.