AppSignal MCP Server
OfficialProvides access to AppSignal's monitoring data, metrics, and tools, enabling AI agents to interact with performance monitoring information from AppSignal.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@AppSignal MCP Servershow me error rates for my production app in the last hour"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
AppSignal MCP Server
This is the official AppSignal MCP server. AppSignal MCP gives your AI editor direct access to your monitoring data: errors, traces, logs, metrics, and dashboards.
AppSignal MCP is a public HTTP endpoint at https://appsignal.com/api/mcp. Connect your agent to that endpoint directly. You don't need this repository to use AppSignal MCP.
This repo is unmaintained. Please use the AppSignal MCP endpoint directly.
This repository holds a legacy stdio proxy, published as the appsignal/mcp Docker image and the @appsignal/mcp npm package. It still works, but it is no longer the recommended way to connect and we do not plan to develop it further. Connect to https://appsignal.com/api/mcp instead. See Legacy stdio proxy if your setup needs it.
Read the full MCP reference on the AppSignal official documentation.
Join our Discord community to help shape this MCP implementation. Feature requests are welcome!
Connect to AppSignal MCP
The endpoint authenticates with OAuth, so there is no server to install or run. In Claude Code:
claude mcp add --transport http appsignal https://appsignal.com/api/mcpClaude Code starts the browser sign-in the first time an AppSignal tool is used. Run claude mcp list to confirm it shows as connected.
The setup guide covers the Claude app, Cursor, Devin, Zed, VS Code, GitHub Copilot CLI, Gemini CLI, and OpenAI Codex. To authenticate with an MCP token instead of OAuth, add an Authorization: Bearer <YOUR_MCP_TOKEN> header. The setup guide shows the exact form for each editor.
Related MCP server: mcp-clickup
What you can access
The endpoint exposes read and write tools across seven areas: error incidents, performance, anomaly detection, logging, metrics, dashboards, and app discovery. With an MCP token, you can set each area to read, write, or disabled. With OAuth, AppSignal exposes all read and write tools at once.
For the full list of tools, parameters, and example prompts, see the MCP tool reference.
Legacy stdio proxy
The Docker image and npm package in this repository wrap the HTTP endpoint in a stdio transport. Use them only for agents that cannot speak HTTP MCP, or in environments that restrict outbound traffic.
This path is obsolete and we do not recommend it. It keeps working, but all new work goes into the HTTP endpoint.
The proxy authenticates with an MCP token, so it exposes the toolsets that token allows. A token created before a tool shipped does not include that tool unless you set the token to expose new tools automatically. Connect to https://appsignal.com/api/mcp over OAuth for all read and write tools, or see Missing any tools?.
Requirements
Docker
An AppSignal account and an AppSignal MCP token
Run the proxy
Pull the Docker image:
docker pull appsignal/mcp:latestThe image reads two environment variables: APPSIGNAL_API_KEY (required, your MCP token) and APPSIGNAL_ENDPOINT (optional, defaults to https://appsignal.com/api/mcp).
In Claude Code:
claude mcp add appsignal -e APPSIGNAL_API_KEY=your-mcp-token -- docker run -i --rm -e APPSIGNAL_API_KEY appsignal/mcpFor other agents, register the same docker run command as a stdio MCP server. Most agents use this shape, but check your agent's own MCP documentation for its exact schema:
{
"mcpServers": {
"appsignal": {
"command": "docker",
"args": ["run", "-i", "--rm", "-e", "APPSIGNAL_API_KEY", "appsignal/mcp"],
"env": {
"APPSIGNAL_API_KEY": "your-mcp-token"
}
}
}
}Development
This repository contains the stdio proxy only. The tools themselves live in the AppSignal application, so new tools and endpoints reach the proxy without a change here.
To work on the proxy:
Start the TypeScript compiler in watch mode:
npm run watchRun tests:
npm testUse the MCP inspector for debugging:
npm run inspector
Change management
Every change that will results in a new version to be released, requires a changeset. Changesets are small Markdown file that describe the change for the end-user. The changeset's frontmatter describes the type of change (new feature, bug fix, etc.) and the version bump (major, minor, or patch).
Use Mono's changeset CLI to generate a new changeset file. Commit the changeset file and include it in your Pull Requests.
mono changeset addPublishing
Install Mono, the tool used for release management.
git pull # Ensure you have the latest version
mono publish # Publish a new versionContributing
Thinking of contributing to our project? Awesome! 🚀
Please follow our Contributing guide in our documentation and follow our Code of Conduct.
Also, we would be very happy to send you Stroopwafles. Have look at everyone we send a package to so far on our Stroopwafles page.
Support
Join our Discord community to chat with other developers and the AppSignal team
Contact us to speak directly with the engineers working on AppSignal. They will help you get set up, tweak your code and make sure you get the most out of using AppSignal.
License
This project is licensed under the MIT License - see the LICENSE file for details.
About AppSignal
AppSignal provides real-time performance monitoring for your web applications. Track errors, monitor performance, measure client-side metrics, and receive alerts when things go wrong.
This server cannot be deployed
Maintenance
Related MCP Connectors
The official MCP Server for the Mux API
Related MCP Servers
- AlicenseNot gradedqualityNot gradedmaintenanceMCP server for Huntress API integration1MIT
- MIT
- Apache 2.0
- MIT