{"id":"win-mcp-server","name":"win-mcp-server","homepage":"https://pypi.org/project/win-mcp-server/","repo_url":"https://github.com/rorymcmahon/win-mcp-server","category":"infrastructure","subcategories":[],"tags":["mcp","winrm","windows","powershell","remote-management","infrastructure-automation","agent-tools"],"what_it_does":"Provides an MCP server that lets AI agents interact with Windows hosts via WinRM, including interactive credential setup and remote PowerShell execution plus basic system/service/disk information tools.","use_cases":["Remote administration and troubleshooting of Windows servers from an AI agent","Automated health checks (system info, running services, disk space)","Running controlled PowerShell commands on managed Windows hosts","Reducing manual ops work for infrastructure teams using MCP-capable agent tooling"],"not_for":["Running untrusted or highly privileged PowerShell on unmanaged hosts","Environments that cannot use WinRM or that require strict network-level restrictions beyond WinRM","Multi-tenant scenarios where credential isolation/auditing must be extremely strong"],"best_when":"You have a controlled Windows fleet with WinRM enabled and you want MCP tool-based remote management with credential caching/clearing on the client side.","avoid_when":"You cannot guarantee host trust, least-privilege credentials, or you need a standard web API/SDK instead of an MCP tool server.","alternatives":["Windows management via Ansible (winrm)","Direct WinRM tooling (e.g., pywinrm scripts) integrated into your own agent platform","Other MCP servers that expose safe, pre-scoped Windows operations (if available)"],"af_score":58.8,"security_score":58.8,"reliability_score":31.2,"package_type":"mcp_server","discovery_source":["pypi"],"priority":"low","status":"evaluated","version_evaluated":null,"last_evaluated":"2026-04-04T21:39:09.640102+00:00","interface":{"has_rest_api":false,"has_graphql":false,"has_grpc":false,"has_mcp_server":true,"mcp_server_url":null,"has_sdk":false,"sdk_languages":[],"openapi_spec_url":null,"webhooks":false},"auth":{"methods":["macOS Keychain credential storage with TouchID (interactive setup on first use)","WinRM authentication using domain credentials (NTLM over HTTP; HTTPS configurable per README)"],"oauth":false,"scopes":false,"notes":"Authentication is described as domain credentials stored in macOS Keychain with TouchID gating, with WinRM using NTLM over HTTP (HTTPS optionally configurable). No agent-scoped authorization model (OAuth scopes) is described."},"pricing":{"model":null,"free_tier_exists":false,"free_tier_limits":null,"paid_tiers":[],"requires_credit_card":false,"estimated_workload_costs":null,"notes":"No pricing information in provided content (GitHub metadata shows small OSS project)."},"requirements":{"requires_signup":false,"requires_credit_card":false,"domain_verification":false,"data_residency":[],"compliance":[],"min_contract":null},"agent_readiness":{"af_score":58.8,"security_score":58.8,"reliability_score":31.2,"mcp_server_quality":78.0,"documentation_accuracy":72.0,"error_message_quality":null,"error_message_notes":null,"auth_complexity":55.0,"rate_limit_clarity":0.0,"tls_enforcement":55.0,"auth_strength":75.0,"scope_granularity":20.0,"dependency_hygiene":60.0,"secret_handling":80.0,"security_notes":"Security claims are described in the README (Keychain + TouchID, hidden password input, credentials cleared from memory, 4-hour expiration, transport security via NTLM over HTTP with HTTPS configurable). However, the README does not provide detailed threat-model, least-privilege guidance, audit/logging details, or how secrets are stored/cleared for all code paths. Execute_powershell exposes powerful capabilities; strong external constraints/allowlists are important.","uptime_documented":0.0,"version_stability":45.0,"breaking_changes_history":30.0,"error_recovery":50.0,"idempotency_support":"false","idempotency_notes":"No explicit guidance on idempotency for execute_powershell or other actions; execute_powershell is inherently non-idempotent depending on the command.","pagination_style":"none","retry_guidance_documented":false,"known_agent_gotchas":["Executing arbitrary PowerShell is dangerous: agents may run destructive or state-changing commands unless you constrain usage.","Credential setup is interactive first use; automated/headless deployment may require additional handling not described.","WinRM transport/security depends on network configuration (HTTP/HTTPS and NTLM settings); agents may fail if hosts are not configured appropriately.","Tool outputs include stdout/stderr and exit codes; agents must parse these reliably rather than assuming structured JSON for all commands."]}}