Seq MCP Server

MCP Server for Seq's API endpoints for interacting with your logging and monitoring system. This server provides comprehensive access to Seq's API features through the Model Context Protocol.

Features

Tools

Signals Management

• `get_signals` - Fetch signals with filtering options
  • Filter by owner ID
  • Filter shared/private signals
  • Support for partial matches

Event Management

• `get_events` - Retrieve events with extensive filtering options
  • Filter by signal IDs
  • Custom filter expressions
  • Configurable event count (max 50)
  • Flexible time range options
  • Date range filtering

Alert Management

• `get_alert_state` - Retrieve the current state of alerts

Resources

Signals Listing

• `signals` - List all shared signals with detailed information
  • Signal ID
  • Title
  • Description
  • Sharing status
  • Owner information

Configuration

The server requires the following environment variables:

• `SEQ_BASE_URL` (optional): Your Seq server URL (defaults to 'http://localhost:8080')
• `SEQ_API_KEY` (required): Your Seq API key

seq-ops Skill

This repo includes a Claude skill for Seq log analysis and incident investigation. Once the MCP server is connected, the skill guides Claude through structured investigations — health checks, incident triage, post-deployment monitoring — and produces consistent, actionable reports.

Install the skill:

# From the releases page, download seq-ops.skill, then:
claude skill install seq-ops.skill

Or install directly from this repo:

npx skills add ahmad2x4/mcp-server-seq

The skill automatically activates when you ask things like "morning health check", "something is broken in prod", "check the logs for timeouts", or "investigate the PPSR integration".

Usage with Claude Code

Run the following command to add the server:

claude mcp add --transport stdio \
  --env SEQ_BASE_URL=http://localhost:5341 \
  --env SEQ_API_KEY=your-api-key \
  seq -- npx -y mcp-seq

To share the configuration with your team, use the `--scope project` flag (saves to `.mcp.json`):

claude mcp add --transport stdio --scope project \
  --env SEQ_BASE_URL=http://localhost:5341 \
  --env SEQ_API_KEY=your-api-key \
  seq -- npx -y mcp-seq

> Windows users: Replace `npx -y mcp-seq` with `cmd /c npx -y mcp-seq` to avoid "Connection closed" errors.

Verify the server is connected by running `/mcp` inside Claude Code.

Usage with Claude Desktop

Add to your `claude_desktop_config.json`:

{
  "mcpServers": {
    "seq": {
      "command": "npx",
      "args": ["-y", "mcp-seq"],
      "env": {
        "SEQ_BASE_URL": "your-seq-url",
        "SEQ_API_KEY": "your-api-key"
      }
    }
  }
}

Development

Install dependencies:

npm install

Build the server:

npm run build

For development with auto-rebuild:

npm run dev

Run tests:

npm run test-script

Time Range Options

The `get-events` tool supports the following time range options:

• `1m` - Last minute
• `15m` - Last 15 minutes
• `30m` - Last 30 minutes
• `1h` - Last hour
• `2h` - Last 2 hours
• `6h` - Last 6 hours
• `12h` - Last 12 hours
• `1d` - Last day
• `7d` - Last 7 days
• `14d` - Last 14 days
• `30d` - Last 30 days

Installation

Install globally via npm:

npm install -g mcp-seq

Or use directly with `npx` — no installation required (as shown in the configuration examples above).

Debugging

Since MCP servers communicate over stdio, debugging can be challenging. The server implements proper error handling and logging for all operations. You can run the test script to verify functionality:

npm run test-script

Type Safety

The server implements comprehensive type safety using:

• TypeScript for static type checking
• Zod schema validation for runtime type checking
• Proper error handling and response formatting