windows-host-mcp (thnak/windows-host-mcp) is an MCP server listed on the M8ven Trust Index. It scores 60 out of 100, grade C. It declares 20 tools. No publisher has claimed this listing.

C
Caution
60/100

windows-host-mcp

An MCP server that runs on Linux and gives an MCP agent (e.g. Claude Code) SSH access to a Windows machine — to build, test, run commands, transfer files, and drive long-running or interactive jobs.

Caution. Specific findings reduced this grade. They are listed on the page. Grades reflect the full trust pyramid: code, verification depth, and reputation. New projects cap at C until adoption is earned.

How we verified

Code Verified⚡ Live Monitored: not connected

Verified is a snapshot. Live keeps it current, and builds your track record.

⚡ Connect GitHub → continuous verification on every pushwhy connect →

Who stands behind it

thnak

Source: Glama

Is this your MCP?

Claim it to get a verified publisher badge, a free copy of our full audit findings, and direct contact for any high-priority issues we find. Or connect your repo for our deepest verification, Live Monitored: read-only, revoke anytime. What we access →

Install from

The grade above is for the source repository. Registries can serve a different version, so we mark the ones we were not able to read.

// key findings
⚠️
Known vulnerabilities in dependencies: 3 high
Affects packages this MCP installs at runtime. Upgrade or remove the affected dependency.
No credential exfiltration, no sensitive file access, no obfuscation
Static analysis found nothing flowing your secrets to unexpected places.
Open source with a license and README
Anyone can audit the code, the license is declared, and the publisher documents what it does.
// tools this server exposes20 tools

These names and descriptions are the publisher's own, read from the source code. We print them as written. Our assessment is the findings above, not this list.

upload_file

Upload a file from this Linux machine to a Windows host over SFTP. Remote paths may use forward or back slashes.

download_file

Download a file from a Windows host to this Linux machine over SFTP — e.g. a build artifact or log.

list_remote_dir

List the contents of a directory on a Windows host. Large listings are capped inline and the full listing spilled to a temp file (.whmcp-tmp/).

sync_dir

Recursively copy a directory between this machine and a Windows host. direction 'up' pushes local→remote (e.g. source), 'down' pulls remote→local (e.g. artifacts).

batch_upload
batch_download
list_hosts

List every configured Windows host with its capability labels (e.g. cuda, rocm), which one is active, and whether it is currently connected. Use this to pick a device, then switch with use_host.

reconnect

Drop and re-establish the SSH connection to a host (defaults to the active host). Use this after changing the Windows account on the host side — e.g. adding it to a group like docker-users, or changing environment/PATH — since an existing SSH session won't pick those up until it reconnects.

use_host

Change which host subsequent commands target by default (the 'device switch'). Select by name, or by a capability label that matches exactly one host (e.g. label 'cuda'). Per-host session state (cwd/env) and jobs are preserved across switches.

start_job

Start a command that may run for a long time (e.g. a build or test suite) without waiting for it to finish. Runs on the active host unless host is given. Returns a jobId (unique across hosts). Poll output with get_job_output, feed stdin with send_job_input, and stop it with cancel_job. Set pty:true

get_job_output

Fetch output from a running or finished job (any host — the jobId identifies it). Pass sinceOffset (the nextOffset from the previous call) to get only new output. Includes status and exit code once the job finishes.

send_job_input

Write input to a running job's stdin — e.g. answer a prompt or feed a REPL. A newline is appended by default.

cancel_job

Stop a running job. Sends a signal and closes the channel. Note: signal delivery is unreliable on Windows OpenSSH; for stubborn processes run a taskkill via run_command.

list_jobs

List jobs with their host, status, and exit codes. Defaults to all hosts; pass host to filter.

get_host_info

One-shot discovery of a Windows host so you don't have to guess: OS version, hostname/user, PATH, and which common tools (git, node, python, dotnet, cmake, docker, nvcc, ...) are installed and their versions. Also reports whether the current session cwd is valid. Run this before a build to avoid cal

check_tools

Given a list of executables, report which are on PATH and their versions. Use this before running a command that depends on a toolchain, instead of discovering it's missing when the command fails.

test_connection

Verify the SSH connection to a host and return its hostname and current user. Defaults to the active host. Use this first to confirm setup.

run_command

Run a one-shot command on a Windows host and wait for it to finish. Returns stdout, stderr, and exit code. Defaults to the active host and its session shell/cwd/env; use start_job instead for long-running or interactive commands.

get_session

Return the default working directory, environment variables, and shell for a host (each host keeps its own). Defaults to the active host.

set_session

Update the default working directory, environment variables, and/or shell used by subsequent run_command / start_job calls on a host. env is merged into the existing defaults (pass null as a value to delete a var). A new cwd is verified to exist on the host before it is accepted, so a typo fails her

// known CVEs in dependencies3 high

Disclosed vulnerabilities in this server's declared npm dependencies (via OSV). Whether each is reachable depends on the installed versions.

high@modelcontextprotocol/sdk@1.12.0GHSA-345p-7cg4-v4c7

@modelcontextprotocol/sdk has cross-client data leak via shared server/transport instance reuse

high@modelcontextprotocol/sdk@1.12.0GHSA-8r9q-7v3j-jr4g

Anthropic's MCP TypeScript SDK has a ReDoS vulnerability

high@modelcontextprotocol/sdk@1.12.0GHSA-w48q-cv73-mx4w

Model Context Protocol (MCP) TypeScript SDK does not enable DNS rebinding protection by default

Depend on this server? Get alerted when its CVEs change.Watch this server free →
// environment variables
To run this server yourself, you supply these values. They go in your own MCP client configuration and stay on your machine. The secret label means the value is sensitive, not that the server mishandles it.
configWINDOWS_HOSTS_CONFIGPoint the server at a JSON file with (see [hosts.example.json](./hosts.example.json)). A single host can still be configured with the flat WINDOWS_SSH_ env vars instead.
configWINDOWS_SSH_DEFAULT_CWD"": "C:\\build"
configWINDOWS_SSH_DEFAULT_SHELLenv =powershell \
configWINDOWS_SSH_HOSTenv =192.168.1.50 \
configWINDOWS_SSH_HOST_NAMEno default Display name for the single env-configured host.
configWINDOWS_SSH_KEY_PASSPHRASEno — Passphrase for the private key, if any.
configWINDOWS_SSH_KEY_PATHenv =/home/you/.ssh/windows_host_ed25519 \
configWINDOWS_SSH_LABELSno — Comma-separated capability labels for the single host.
configWINDOWS_SSH_PORTno 22 SSH port.
configWINDOWS_SSH_READY_TIMEOUT_MSno 20000 SSH handshake timeout.
configWINDOWS_SSH_USERenv =builder \
configXDG_CONFIG_HOME
// quality suggestions

Tool annotations

No tools have read-only/destructive annotations

Add readOnlyHint or destructiveHint annotations to every tool so hosts can warn users before invoking.

All four hints declared on every tool

20/20 tools missing one or more hints — upload_file (missing: readOnlyHint, destructiveHint, idempotentHint, openWorldHint); download_file (missing: readOnlyHint, destructiveHint, idempotentHint, openWorldHint); list_remote_dir (missing: readOnlyHint, destructiveHint, idempotentHint, openWorldHint), +17 more. OpenAI's directory rejects tools where any of the four hints are missing or non-boolean.

For every tool, set all four hints (readOnlyHint, destructiveHint, idempotentHint, openWorldHint) to explicit true/false values that match the handler’s actual behaviour.

Tool inputs are validated

17/20 tool handlers declare input schemas (85%)

Declare an inputSchema with zod/joi/yup on every tool definition.

Tests exist

No test files found

Add tests that exercise each declared tool.

Shell command execution

3 child_process calls — runs shell commands

Prefer library functions over shell-outs. If you must shell out, ensure all inputs are properly escaped.

Production dependencies are patched

0 critical, 3 high severity in production deps — @modelcontextprotocol/sdk@1.12.0 (high), @modelcontextprotocol/sdk@1.12.0 (high)

Run npm audit fix, or upgrade the affected packages to a non-vulnerable version.

Claim the listing to review these findings one by one and send us a correction where you disagree, straight to the team. Claiming also means we tell you when the grade moves, and reach you first if we find anything urgent.

// full audit trail
The findings above are the summary. The full trail, every check we ran, each deduction, the network hosts observed and the dependency advisories, goes to verified publishers, along with an alert whenever a new one lands. Verified publishers can also review each finding and dispute it in one click. Publisher corrections have sharpened several of our checks this month, because the maintainer knows the codebase better than any scanner.
// improvement guidance — verified publishers only
We have 6 concrete improvements we can share with the publisher of this MCP. Each comes with specific guidance to raise the trust score.
// embed badge in your README
[![M8ven Verified](https://m8ven.ai/badge/mcp/thnak/windows-host-mcp?variant=verified)](https://m8ven.ai/mcp/thnak/windows-host-mcp)
Shows verification status without the grade. Want the grade badge instead? Remove ?variant=verified from the URL.
commit: c9aa14e5f91265ed794d20d87d58e3fa8b255fb2
code hash: 369516ac63149a3e3e0a970b5ec688b8902d05dd402e823f17b332e48bd9107d
view raw JSON →
Check MCPs from inside your assistant
Tool Check · MCP

Vetting this one by hand? Tool Check is an MCP that scores other MCPs. Add it once and ask Claude, ChatGPT, or any MCP client to grade a server, surface CVEs, check the publisher, and suggest safer alternatives — before you install.

https://m8ven.ai/api/mcp/tool-check
check_toolsearch_toolscompare_toolsrecommend_alternativescheck_publisherreport_concern
How to add it →Free · no account needed · works in any MCP client