Windows Operations MCP
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
No arguments | |||
Capabilities
Features and capabilities supported by this server
| Capability | Details |
|---|---|
| tools | {
"listChanged": true
} |
| logging | {} |
| prompts | {
"listChanged": false
} |
| resources | {
"subscribe": false,
"listChanged": false
} |
| extensions | {
"io.modelcontextprotocol/ui": {}
} |
| experimental | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| command_executionA | Execute Windows commands with reliable output capture and agentic sampling. RATIONALE: Consolidates PowerShell and CMD execution into a single async portmanteau. Uses asyncio.to_thread() to ensure the MCP event loop is never blocked. Integrates with FastMCP 3.2 Context for real-time telemetry and sampling. PATTERNS:
Args: action: Execution environment ("powershell" or "cmd"). command: The command string to execute. working_directory: Optional CWD for the command. timeout_seconds: Hard timeout (1-300s, default: 30). max_output_size: Truncation limit for high-volume logs. ctx: FastMCP Context for telemetry and sampling (injected). Examples: - command_execution(action="powershell", command="Get-Service | Where-Object Status -eq 'Running'") - command_execution(action="cmd", command="dir /s /b *.log") |
| archive_managementC | Perform Windows archive management operations with agentic telemetry. RATIONALE: Standardizes archive handling for automation and forensics. Provides specialized support for Windows-native CAB (.cab) files via expand.exe. Args: action: The archive operation to perform. path: Path to the archive file. target_dir: Directory to extract to (for "extract" and "expand_cab"). source_files: List of files to compress (for "create" and "add_file"). archive_type: Format to use for creation. ctx: FastMCP Context for telemetry and sampling. |
| json_operationsB | Perform specialized JSON data operations with agentic telemetry. RATIONALE: Agents often need to "patch" existing configs or extract JSON from unstructured logs. This portmanteau provides the specialized logic required for high-level data surgery. Args: action: The JSON operation to perform. path: File path for read/write/patch operations. data: Data to write or merge (for "write"/"patch"). text: Raw text to extract JSON from (for "extract_from_text"). indent: Indentation for formatting. ctx: FastMCP Context for telemetry and sampling. |
| process_managementC | Perform process management operations with comprehensive monitoring and agentic telemetry. Returns: Union[Dict, ToolResult] depending on action and availability of prefab-ui. |
| windows_servicesB | Perform Windows service operations with comprehensive error handling and agentic telemetry. Returns: Union[Dict, ToolResult] depending on action and availability of prefab-ui. |
| system_managementA | Perform system management operations with comprehensive error handling and agentic telemetry. RATIONALE: Consolidates system info, health checks, and network connectivity into a single portmanteau. Integrates with FastMCP 3.2 Context for real-time progress reporting and LLM-in-the-loop diagnostics. Args: action: The system operation to perform. detailed: Include additional technical details (default: False). host: Target hostname for port testing. port: Target port for connectivity verification. timeout_seconds: Connection timeout (default: 5s). category: Help category filter. ctx: FastMCP Context for telemetry and sampling (injected). Examples: - system_management(action="health", detailed=True) - system_management(action="test_port", host="8.8.8.8", port=53) |
| windows_event_logsA | Perform Windows Event Log operations with comprehensive error handling and agentic telemetry. RATIONALE: Consolidates querying, clearing, and exporting logs into a single async portmanteau. Uses asyncio.to_thread for blocking pywin32 calls to maintain MCP responsiveness. Args: action: The log operation to perform. log_name: Name of the event log (Application, System, Security, etc.). max_events: Maximum events to return (for "query"). time_range_hours: Lookback window (for "query"). event_id: Filter by specific Event ID. output_path: Destination for exported logs (for "export"). ctx: FastMCP Context for telemetry and sampling. |
| windows_performanceB | Perform Windows Performance monitoring with comprehensive error handling and agentic telemetry. RATIONALE: Consolidates system-wide, process-specific, and low-level counter monitoring into a single portmanteau. Integrates with FastMCP 3.2 Context for real-time progress reporting and LLM-in-the-loop diagnostics. Args: action: The performance operation to perform. pid: Specific process ID to monitor (for "process"). include_network: Include network I/O stats. duration_seconds: Interval for CPU sampling (default: 1s). ctx: FastMCP Context for telemetry and sampling. |
| windows_permissionsA | Perform Windows Permission (ACL) operations with comprehensive error handling and agentic telemetry. RATIONALE: Consolidates ACL viewing, granting, revoking, and inheritance management into a single portmanteau. Uses 'icacls.exe' for industrial reliability on Windows. Args: action: The permission operation to perform. path: Target file or directory path. user: Target user or group (required for grant/revoke). permission: Permission level (F=Full, M=Modify, RX=Read/Exec, R=Read, W=Write). enable_inheritance: Whether to enable or disable inheritance (for "inheritance"). ctx: FastMCP Context for telemetry and sampling. |
| windows_accountsC | Perform Windows local account and group management with agentic telemetry. RATIONALE: Enables autonomous identity and access management (IAM) on local Windows systems. Uses 'net.exe' for industrial reliability and broad OS compatibility. Args: action: The account operation to perform. user: Target username. password: New password (for add_user or set_password). group: Target local group name. group_action: Action to perform on the group (target user is 'user'). ctx: FastMCP Context for telemetry and sampling. |
| windows_automationC | Perform Windows Automation operations: Scheduled Tasks and WMI Queries. RATIONALE: Enables autonomous system orchestration and deep environment forensics. Uses 'schtasks.exe' and 'wmic.exe' for industrial reliability. Args: action: The automation operation to perform. task_name: Unique name for the scheduled task. task_path: Executable path for the scheduled task (for "create_task"). schedule: Task frequency. start_time: Task start time (format HH:mm). wmi_class: Target WMI class (e.g., 'Win32_Processor', 'Win32_BIOS'). wmi_namespace: WMI namespace for the query. ctx: FastMCP Context for telemetry and sampling. |
| windows_networkB | Perform Windows networking and firewall operations with specialized orchestration. RATIONALE: Consolidates complex 'netsh' and PowerShell networking commands into a single portmanteau. Enables autonomous security auditing and perimeter hardening. Args: action: The networking operation to perform. rule_name: Name of the firewall rule. rule_dir: Direction of the traffic (in/out). rule_action: Action (allow/block). rule_program: Path to the executable (for firewall_add). rule_port: Local port number (for firewall_add). ctx: FastMCP Context for telemetry and sampling. |
| windows_appsB | Manage modern Windows AppX and Store packages with PowerShell orchestration. RATIONALE: Modern Windows applications (AppX) cannot be managed via traditional 'net.exe' or basic registry surgery. This tool uses PowerShell to enable autonomous bloatware removal and package auditing. Args: action: The apps operation to perform. name_filter: Filter for listing (e.g. 'Xbox', 'Bing'). package_name: Exact package name to uninstall. all_users: Perform action for all users (requires elevation). ctx: FastMCP Context for telemetry and sampling. |
| agentic_system_hardeningB | Execute an autonomous system hardening mission with SEP-1577 Sampling. RATIONALE: This orchestrator coordinates multiple specialized tools to reach a security baseline. It uses 'Reasoning-First' sampling to identify vulnerabilities before action. Args: target: The subsystem to harden. dry_run: If True, only audit and recommend (default: True). ctx: FastMCP Context for telemetry and sampling. |
| autonomous_troubleshooterA | Diagnose WHY a Windows operation failed (Permissions vs Process vs Registry). Uses a 3-phase approach:
Args: operation_failure: Description of the failure to investigate. ctx: FastMCP Context for telemetry and sampling (required). |
| system_health_cardA | Display a rich system health card with CPU, memory, and disk IO stats. Returns a Prefab UI card in capable MCP hosts; plain text fallback otherwise. |
| process_list_cardA | Display a rich card listing running processes, optionally filtered by name. Returns a Prefab UI card in capable MCP hosts; plain text fallback otherwise. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
| registry_hardening_wizard | Guide for identifying and fixing insecure registry keys in a given hive. |
| powershell_agent_scaffold | Generate a robust, error-tolerant PowerShell script scaffold for a given Windows task. |
| system_account_audit | Review local user accounts and group memberships for security and privilege alignment. |
| data_surgery_forensics | Guide for using JSON and Archive tools to collect and analyze system configuration artifacts. |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
| get_llms_txt | LLM-friendly summary of all tools and capabilities in this server. |
| get_llms_full_txt | Full LLM corpus for windows-operations-mcp (llms-full.txt). |
| get_expert_skill_legacy | [Legacy] SOTA Windows Expert skill instructions (use skill://windows-expert/SKILL.md instead). |
| Prefab Renderer | |
| windows-expert/SKILL.md | 🛠️ Skill: Windows Native Hardening & Data Surgery |
| windows-expert/_manifest | File listing for windows-expert |
Latest Blog Posts
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/sandraschi/windows-operations-mcp'
If you have feedback or need assistance with the MCP directory API, please join our Discord server