DevOps / SRE5 MCP serversMCP 2025-06-18 Stable

DevOps / SRE Loadout

5 MCP servers for incidents, deploys, and on-call sanity

Last verified:

When the pager goes off, you need to read code, check the cluster, look at errors, and acknowledge in PagerDuty — fast. This loadout puts all of that one chat away.

Install DevOps / SRE Loadout

Paste into ~/Library/Application Support/Claude/claude_desktop_config.json
{
  "mcpServers": {
    "filesystem": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-filesystem", "/Users/you/infra"]
    },
    "github": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-github"],
      "env": { "GITHUB_PERSONAL_ACCESS_TOKEN": "ghp_..." }
    },
    "kubernetes": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-kubernetes"]
    },
    "sentry": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-sentry"],
      "env": { "SENTRY_AUTH_TOKEN": "..." }
    },
    "pagerduty": {
      "command": "npx",
      "args": ["-y", "pagerduty-mcp"],
      "env": { "PAGERDUTY_API_KEY": "..." }
    }
  }
}

After pasting, fully restart your MCP client. Replace placeholders (paths, API keys) with your own values.

What's in this loadout

  • Filesystem MCP Server
    coremcp

    Read infra-as-code repos, runbooks, and incident postmortems.

  • GitHub MCP Server
    coremcp

    Find the commit that introduced a regression, open a hotfix PR, link it in the incident channel.

  • Kubernetes MCP Server
    coremcp

    List unhealthy pods, describe deployments, get logs without `kubectl` muscle memory.

  • Sentry MCP
    coremcp

    Correlate errors to deploys and surface what regressed.

  • PagerDuty MCP
    optionalmcp

    Acknowledge incidents and read the on-call schedule from the same chat.

What you can do with it

  • Triage an alert

    A pagerduty alert just fired for "checkout-api error rate > 5%". Pull the last 50 errors from sentry, identify the deploy on github that introduced the regression, list unhealthy pods in kubernetes, and propose either a rollback or a hotfix.

  • Postmortem draft

    For pagerduty incident #4821, gather: timeline of the alert, the github commits in the rollback window, sentry error counts before/after, and unhealthy pods during the window. Draft a postmortem in /Users/me/runbooks/postmortems/.

  • Healthy state check

    List all kubernetes deployments in the "production" namespace where readiness probes are failing. Cross-reference with sentry errors in the last hour.

Client compatibility

  • Claude Desktop 0.7.4
    Tested ✓
  • Cursor 0.42+
    Tested ✓
  • Windsurf 1.4
    Tested ✓
  • Continue 0.9
    Untested

Variants

Heavy version

Adds Datadog for metrics + traces and Terraform for infra changes.

Adds: datadog-mcp, terraform-mcp

Lite version

Filesystem + Kubernetes + Sentry. The bare minimum during an incident.

Removes: mcp-server-github, pagerduty-mcp

FAQ — DevOps / SRE Loadout

  • What is the DevOps / SRE Loadout?

    When the pager goes off, you need to read code, check the cluster, look at errors, and acknowledge in PagerDuty — fast. This loadout puts all of that one chat away. It includes 5 MCP servers: Filesystem MCP Server, GitHub MCP Server, Kubernetes MCP Server, Sentry MCP, PagerDuty MCP.

  • Does this loadout work with Cursor and Windsurf?

    Yes. The config block above is one JSON object that drops into Claude Desktop's, Cursor's, or Windsurf's MCP config. Each client uses the same mcpServers schema; only the file path differs.

  • How long does setup take?

    About 3 minutes the first time: copy the config, paste it into your client's MCP configuration file, replace placeholder API keys with your own, and restart the client.

  • Can I customise this loadout?

    Yes. Use the Lite version if you want fewer servers, or the Heavy version to add 2 more for advanced workflows. You can also remove any individual server by deleting its entry from the JSON.

  • How much does it cost?

    Loadout itself is free. Most listed servers are open-source and free to run. Servers that talk to paid SaaS (Stripe, SendGrid, Linear, etc.) follow that vendor's pricing.

  • When was this loadout last verified?

    Last verified on 2026-04-24. We re-test featured loadouts at least monthly and update the config when MCP servers ship breaking changes.