Overview

The Seq MCP Server enables AI assistants to search, stream, and analyze structured log events from Datalust Seq, a centralized structured log server. Built as a pure MCP stdio-based service in C# on .NET 9, this server communicates via JSON-RPC and provides tools for querying log events, waiting for live events, and listing signals — bringing observability data directly into AI conversations.

Community Server:

Developed by willibrandon. Available as a .NET global tool on NuGet or via Docker.

Key Features

🔍

Flexible Log Search

Search Seq events with powerful filter expressions — by level, application, error count, or custom properties.

🔴

Live Event Streaming

Capture and analyze live events from Seq with a configurable timeout for real-time monitoring.

🏷️

Signal Integration

List and inspect Seq signals to understand saved queries and alert conditions.

🔧

Workspace Support

Query specific workspaces with workspace-scoped API keys for multi-environment deployments.

Available Tools

Quick Reference

Tool	Purpose	Category
`SeqSearch`	Search Seq events with filter expressions	Read
`SeqWaitForEvents`	Wait for and capture live events with timeout	Stream
`SignalList`	List available signals in Seq	Read

Detailed Usage

SeqSearch▶

Search Seq events using filter expressions. Supports property-based filtering and workspace-scoped queries.

use_mcp_tool({
  server_name: "seq",
  tool_name: "SeqSearch",
  arguments: {
    filter: "@Level = 'Error'",
    count: 50
  }
});

Use empty string "" for all events. Supports filters like Application = "MyApp", @Level = "Error", or free text.

SeqWaitForEvents▶

Wait up to 5 seconds for live events matching a filter. Returns a snapshot of events captured during the wait period.

use_mcp_tool({
  server_name: "seq",
  tool_name: "SeqWaitForEvents",
  arguments: {
    filter: "@Level = 'Warning'",
    count: 10
  }
});

Useful for real-time monitoring. May return empty if no events match during the wait period.

SignalList▶

List available signals from your Seq server. Signals represent saved queries or alert conditions.

use_mcp_tool({
  server_name: "seq",
  tool_name: "SignalList",
  arguments: {}
});

Optionally specify a workspace parameter to filter signals by workspace.

Installation

Install the .NET global tool:

dotnet tool install -g SeqMcpServer

Then configure your MCP client:

{
  "mcpServers": {
    "seq": {
      "command": "seq-mcp-server",
      "env": {
        "SEQ_SERVER_URL": "http://localhost:5341",
        "SEQ_API_KEY": "your-api-key"
      }
    }
  }
}

Requirements:

Requires .NET 9.0 Runtime or SDK. Update with dotnet tool update -g SeqMcpServer.

Configuration

The Seq MCP server uses environment variables for configuration:

Variable	Required	Description
`SEQ_SERVER_URL`	Yes	URL of your Seq server (e.g. `http://localhost:5341`)
`SEQ_API_KEY`	Yes	API key for authenticating with Seq
`SEQ_API_KEY_<WORKSPACE>`	No	Workspace-specific API keys (e.g. `SEQ_API_KEY_PRODUCTION`)

Common Use Cases

1. Error Investigation

Search for recent errors to diagnose application issues:

use_mcp_tool({
  server_name: "seq",
  tool_name: "SeqSearch",
  arguments: {
    filter: "@Level = 'Error'",
    count: 20
  }
});

2. Real-time Monitoring

Watch for warnings as they come in:

use_mcp_tool({
  server_name: "seq",
  tool_name: "SeqWaitForEvents",
  arguments: {
    filter: "@Level = 'Warning'",
    count: 5
  }
});

3. Application-Specific Analysis

Filter events from a specific application:

use_mcp_tool({
  server_name: "seq",
  tool_name: "SeqSearch",
  arguments: {
    filter: "Application = 'WebApp'",
    count: 100
  }
});

Seq MCP Server: Real-time Log Analytics for AI Models