Skip to content
Argo CD MCP Server

Argo CD MCP Server

Released
v0.9.0
Apache-2.0
argocdgitopskubernetesdevopscontinuous-deliveryargoproj

Summary

Argo CD's MCP server from argoproj-labs — list and sync applications, walk the resource tree and read workload logs in natural language.

Features

  • List, create, update, delete and sync Argo CD applications
  • Walk an application's resource tree and managed resources
  • Read workload logs and resource events without leaving the chat
  • List and run Argo CD resource actions
  • Read-only mode strips the mutating tools for production access
  • Stateless mode plus a token registry serves multiple Argo CD instances
  • stdio and HTTP stream transports, with self-signed certificate support

Installation

Set up this MCP server in your favorite AI agent — copy a ready-made configuration below.

Any MCP-compatible agent

Most agents (Claude, Cursor, Windsurf, VS Code, and more) read a standard mcpServers configuration.

  1. Open your agent's MCP configuration file.
  2. Merge the snippet below into it, filling in the environment variables with your own values.
  3. Restart the agent — the "Argo CD MCP Server" tools will be available.
{
  "mcpServers": {
    "argocd-mcp-server": {
      "command": "npx",
      "args": [
        "argocd-mcp@latest",
        "stdio"
      ],
      "env": {
        "ARGOCD_BASE_URL": "https://argocd.example.com",
        "ARGOCD_API_TOKEN": "your-argocd-api-token"
      }
    }
  }
}

Description

An MCP server for Argo CD, maintained under argoproj-labs, the Argo project's own incubator organisation. It exposes the Argo CD API so a GitOps investigation — why is this app out of sync, what does the resource tree look like, what do the pods say — happens in one conversation instead of a browser tab plus three kubectl invocations.

Tools

Applicationslist_applications (with filters), get_application, create_application, update_application, delete_application and sync_application to trigger a sync.

Resourcesget_application_resource_tree for the full tree, get_application_managed_resources, get_application_workload_logs for logs from Pods, Deployments and other workloads, get_resource_events, and get_resource_actions plus run_resource_action to list and execute the actions Argo CD offers on a resource.

Clusters and projectslist_clusters for registered clusters and get_appproject for AppProject detail.

Deployment notes

Both stdio and HTTP stream transports are supported. Two features matter for anything beyond a laptop:

  • Read-only mode removes the mutating tools, which is the sane default for handing an agent access to a production Argo CD.
  • Stateless mode and a token registry let one deployment serve multiple Argo CD instances, with a default instance plus extras resolved per request.

Self-signed certificates are supported, and the docs cover network exposure and the several ways of supplying credentials.

Requirements

Node.js 18 or newer, an Argo CD instance with API access, and an Argo CD API token. Runs with npx argocd-mcp@latest stdio; one-click install links are provided for VS Code and VS Code Insiders, with configuration examples for Cursor and Claude Desktop. Apache-2.0.

Related MCP Servers

MCP: AWS MCP Server

by Amazon Web Services

New

AWS's managed MCP server: call any AWS API, run sandboxed Python against your account, and load AWS agent skills on demand — under IAM, CloudWatch and CloudTrail.

Cloud & Infrastructure
New

Google's official MCP server for driving the gcloud CLI from natural language, with allowlist and denylist controls over which commands an agent may run.

Cloud & Infrastructure
1 views
Featured

Google's open-source MCP server for enterprise databases — one binary that fronts AlloyDB, BigQuery, Spanner, Postgres, MySQL, Oracle, MongoDB, Snowflake and more.

Databases & DataCloud & Infrastructure
4 views
Featured

Datadog's official managed MCP server — query logs, metrics, traces, incidents and dashboards from an AI agent, inside your existing Datadog permissions.

Cloud & Infrastructure
Browse all MCP servers →