MCP SysOperator

A Model Context Protocol (MCP) server for Infrastructure as Code operations. This server allows AI assistants to interact with Ansible and Terraform, execute playbooks and Terraform plans, manage cloud resources, and perform other infrastructure operations directly.
(Project previously known as _mcp-ansible_)

Demo Projects

__All code in demos generated using Claude 3.7 Sonnet (via OpenRouter), Cline, and SysOperator__

- AWS LAMP Stack - All Ansible code to deploy example LAMP stack in AWS
- AWS Terraform LAMP - Terraform and Ansible code to reply a LAMP stack

Features

- Run Ansible Playbooks: Execute Ansible playbooks with support for parameters like inventory, extra vars, tags, and limits
- List Inventory: View hosts and groups from an Ansible inventory file
- Check Syntax: Validate Ansible playbook syntax without execution
- List Tasks: Preview tasks that would be executed by a playbook
- Access Default Inventory: Access the default Ansible inventory file via resource API
- AWS Integration: Manage AWS resources (EC2, S3, VPC, CloudFormation, etc.)
- Terraform Support: Execute Terraform commands (init, plan, apply, destroy, output, etc.)
- tflocal Integration: Test Terraform configurations with LocalStack for local cloud development
- LocalStack Support: Test AWS operations locally using LocalStack without real AWS credentials

Requirements

- Node.js 18 or higher
- npm or yarn
- Ansible installed and in PATH
- @modelcontextprotocol/sdk (installed automatically)
- For AWS operations: AWS CLI and valid credentials
- For LocalStack: LocalStack installed and running, awslocal CLI

Installation

$3

``bash
git clone https://github.com/tarnover/mcp-sysoperator.git
cd mcp-sysoperator
`

$3

`bash
npm install
`

$3

`bash
npm run build
`

$3

Add the Ansible MCP server to your MCP settings configuration file.

For VSCode with Claude extension:
- Edit the file at ~/.config/Code/User/globalStorage/saoudrizwan.claude-dev/settings/cline_mcp_settings.json

For Claude Desktop app:
- macOS: Edit ~/Library/Application Support/Claude/claude_desktop_config.json
- Windows: Edit %APPDATA%\Claude\claude_desktop_config.json
- Linux: Edit ~/.config/Claude/claude_desktop_config.json

Add the following to the mcpServers section:

`json
{
"mcpServers": {
"sysoperator": {
"command": "node",
"args": ["/absolute/path/to/mcp-sysoperator/build/index.js"],
"env": {}
}
}
}
`

Make sure to replace /absolute/path/to/mcp-sysoperator with the actual path to your installation.

Usage Examples

Once installed and configured, the MCP server provides the following tools to the AI assistant:

$3

`

sysoperator
run_playbook

{
"playbook": "/path/to/your/playbook.yml",
"inventory": "/path/to/inventory.ini",
"extraVars": {
"var1": "value1",
"var2": "value2"
},
"tags": "setup,configure",
"limit": "webservers"
}


`

$3

`

sysoperator
list_inventory

{
"inventory": "/path/to/inventory.ini"
}


`

$3

`

sysoperator
check_syntax

{
"playbook": "/path/to/your/playbook.yml"
}


`

$3

`

sysoperator
list_tasks

{
"playbook": "/path/to/your/playbook.yml"
}


`

$3

`

sysoperator
sysoperator://inventory/default

`

$3

`

sysoperator
aws_s3

{
"action": "list_buckets",
"region": "us-east-1"
}


`

$3

`

sysoperator
terraform

{
"action": "init",
"workingDir": "/path/to/terraform/project"
}




sysoperator
terraform

{
"action": "plan",
"workingDir": "/path/to/terraform/project",
"vars": {
"instance_type": "t2.micro",
"region": "us-west-2"
}
}


`

$3

`

sysoperator
terraform

{
"action": "apply",
"workingDir": "/path/to/terraform/project",
"autoApprove": true,
"vars": {
"instance_type": "t2.micro",
"region": "us-west-2"
}
}


`

$3

`

sysoperator
terraform

{
"action": "apply",
"workingDir": "/path/to/terraform/project",
"useLocalstack": true,
"autoApprove": true,
"vars": {
"instance_type": "t2.micro",
"region": "us-west-2"
}
}


`

LocalStack Integration

This project includes integration with LocalStack for testing AWS operations locally without real AWS credentials. The LocalStack integration allows you to:

1. Test Ansible playbooks that use AWS services locally
2. Develop and test AWS operations without incurring AWS costs
3. Run tests without requiring real AWS credentials
4. Validate your infrastructure code before deploying to real AWS

$3

See the LocalStack README for detailed instructions on using the LocalStack integration.

Quick start:

`bash


Install LocalStack and awslocal CLI

pip install localstack awscli-local

Start LocalStack

localstack start

Run the sample playbook

node localstack/run_sample_playbook.mjs
`

Development

$3

`
mcp-sysoperator/
├── src/
│ ├── index.ts # Main entry point
│ └── ansible-mcp-server/ # Will be renamed in filesystem in future updates
│ ├── index.ts # MCP SysOperator server implementation
│ ├── common/ # Common utilities and types
│ │ ├── errors.ts # Error definitions
│ │ ├── types.ts # Type and schema definitions
│ │ ├── utils.ts # Utility functions
│ │ └── version.ts # Version information
│ └── operations/ # Operation handlers
│ ├── ad_hoc.ts # Ansible ad-hoc commands
│ ├── aws.ts # AWS operations
│ ├── inventory.ts # Ansible inventory operations
│ ├── playbooks.ts # Ansible playbook operations
│ ├── terraform.ts # Terraform operations
│ └── vault.ts # Ansible vault operations
├── localstack/ # LocalStack integration
│ ├── README.md # LocalStack documentation
│ ├── sample_playbook.yml # Sample playbook for LocalStack
│ ├── inventory.ini # Sample inventory for LocalStack
│ ├── run_sample_playbook.mjs # Script to run sample playbook
│ └── utils.localstack.ts # Modified utils for LocalStack
├── package.json # Project configuration and dependencies
├── tsconfig.json # TypeScript configuration
└── README.md # Documentation
`

$3

To add new capabilities to the MCP server:

1. Modify src/ansible-mcp-server/index.ts (future: src/sysoperator/index.ts)
2. Add your new tool in the setupToolHandlers method
3. Implement a handler function for your tool in the appropriate operations file
4. Add the schema definition in common/types.ts
5. Rebuild with npm run build`

$3

SysOperator is currently in active development and undergoing extensive testing. It is not recommended for use in production environments at this time. The software may experience breaking changes, incomplete features, or unexpected behavior.

Use at your own risk.

License

MIT License - See LICENSE for details