This project implements a Model Context Protocol (MCP) server that integrates with Prometheus Alertmanager. It serves as a bridge between Claude AI and Alertmanager, allowing Claude to interact with and manage alerts through a standardized interface.
-
Alert Retrieval: Fetches and formats current alerts from Alertmanager with optional filtering capabilities.
-
Alert Details: Provides detailed information about specific alerts when referenced by their fingerprint.
-
Silence Management: Offers tools to create, list, and delete silences that suppress notifications for specific alerts.
-
Alert Grouping: Retrieves alert groups as organized by Alertmanager.
- Built using TypeScript and the MCP SDK
- Communicates with Alertmanager's HTTP API (v2)
- Runs as a stdio-based process compatible with Claude for Desktop
- Implements proper error handling and timeout management
- Provides strongly-typed interfaces for alert and silence data
- Access to a running Prometheus Alertmanager instance
- For Claude for Desktop: Node.js (v18 or newer) installed on your system
The package is available on npm and can be used directly with npx:
npx alertmanager-mcp
For global installation:
npm install -g alertmanager-mcp
alertmanager-mcp
- Configure Claude for Desktop to use the MCP server by editing the configuration file:
{ "mcpServers": { "alertmanager": { "command": "docker", "args": [ "run", "--rm", "-i", "--network=host", "-e", "ALERTMANAGER_URL=http://your-alertmanager-url:9093", "ghcr.io/kaznak/alertmanager-mcp:latest" ], "env": {} } } }
2. Restart Claude for Desktop to load the new configuration.
3. You can now ask Claude to interact with Alertmanager using natural language:
- "Show me current alerts"
- "Filter alerts related to CPU issues"
- "Get details for this alert"
- "Create a silence for this alert for the next 2 hours"
## Available Tools
### get-alerts
Retrieves a list of alerts with optional filtering.
Parameters:
- `filter`: (optional) Filtering query (e.g., `alertname=~'.*CPU.*'`)
- `silenced`: (optional) Include silenced alerts
- `inhibited`: (optional) Include inhibited alerts
- `active`: (optional) Include active alerts (default: true)
### get-alert-details
Gets detailed information about a specific alert.
Parameters:
- `fingerprint`: Alert fingerprint
### create-silence
Creates a silence for alerts matching specified criteria.
Parameters:
- `matchers`: List of matchers for alerts
- `startsAt`: (optional) Silence start time (ISO8601 format, default is current time)
- `endsAt`: Silence end time (ISO8601 format)
- `createdBy`: Username who created the silence
- `comment`: Reason or explanation for the silence
### get-silences
Retrieves a list of silences with optional filtering.
Parameters:
- `filter`: (optional) Filtering query (e.g., `createdBy=~'.*admin.*'`)
### delete-silence
Deletes a silence by ID.
Parameters:
- `silenceId`: ID of the silence to delete
### get-alert-groups
Gets alert groups with optional filtering.
Parameters:
- `active`: (optional) Include active alerts (default: true)
- `silenced`: (optional) Include silenced alerts
- `inhibited`: (optional) Include inhibited alerts
## Extending the Server
This MCP server can be extended with additional features such as:
- Alert trend analysis
- Automatic response suggestions
- Integration with incident management systems
- Custom dashboards for specific alert types
## Resources
- [Model Context Protocol Documentation](https://modelcontextprotocol.io/docs/concepts/architecture)
- [Prometheus Alertmanager Documentation](https://prometheus.io/docs/alerting/latest/alertmanager/)
- [Alertmanager API Reference](https://prometheus.io/docs/alerting/latest/api/)
- [MCP TypeScript SDK](https://github.com/modelcontextprotocol/typescript-sdk)
- [Claude AI Documentation](https://docs.anthropic.com/claude/)