mfa-servicenow-mcp@1.25.49
CAUTION· not startedPublisher claims: "A Model Context Protocol (MCP) server implementation for ServiceNow" Observed in sandbox: no network, touches nothing outside the project, writes no memory.
Coverage
Not started: dependency install failed; the server itself was not observed.- Tested
- the dependency install step · every file read, file written and host contacted by what ran
- Not tested
- server startup · the tool list · tool behavior
What this means for the label: Caution here covers what ran. The server's tools were not exercised with a real credential, so nothing about their behavior is asserted; the credential summary below says what to do with that.
Before you give it your token
- It will ask for
SERVICENOW_API_KEYSERVICENOW_API_KEY_HEADERSERVICENOW_AUTH_TYPESERVICENOW_BROWSER_PASSWORDSERVICENOW_CLIENT_SECRETSERVICENOW_MCP_HTTP_AUTH_TOKENSERVICENOW_PASSWORDSERVICENOW_TOKEN_URLor--http-auth-token/--client-secret/--browser-passwordConfiguration, not credentials: LOG_FILE, MCP_TOOL_PACKAGE, PLAYWRIGHT_BROWSERS_PATH, SERVICENOW_ACTIVE_INSTANCE, SERVICENOW_BROWSER_DEBUG, SERVICENOW_BROWSER_HEADLESS, SERVICENOW_BROWSER_LOGIN_URL, SERVICENOW_BROWSER_PROBE_PATH, SERVICENOW_BROWSER_SESSION_TTL, SERVICENOW_BROWSER_TIMEOUT, SERVICENOW_BROWSER_USER_DATA_DIR, SERVICENOW_BROWSER_USERNAME, SERVICENOW_CLIENT_ID, SERVICENOW_DEBUG, SERVICENOW_INSTANCE_URL, SERVICENOW_MCP_HTTP_ALLOWED_HOSTS, SERVICENOW_MCP_HTTP_HOST, SERVICENOW_MCP_HTTP_PATH, SERVICENOW_MCP_HTTP_PORT, SERVICENOW_MCP_SERVER_NAME, SERVICENOW_MCP_TRANSPORT, SERVICENOW_TIMEOUT, SERVICENOW_USERNAME, TOOL_PACKAGE_CONFIG_PATH- Where it talks to
- no vendor host named in the codeAlso service-now.com, evil.example, opencode.ai — not the publisher's domain; check what for before connecting.
- What we saw when it ran
- no credential files read · no network
- What we could not test
- server startup · the tool list · tool behavior
Connect it with a scoped credential, under the guard, and read the third-party hosts first. The code names service-now.com, evil.example, opencode.ai as well as its own vendor. The guard injects SERVICENOW_API_KEY, SERVICENOW_API_KEY_HEADER, SERVICENOW_AUTH_TYPE, SERVICENOW_BROWSER_PASSWORD, SERVICENOW_CLIENT_SECRET, SERVICENOW_MCP_HTTP_AUTH_TOKEN, SERVICENOW_PASSWORD, SERVICENOW_TOKEN_URL only into requests to the host you bind (and its subdomains) and refuses the rest.
Never paste a credential into a chat, a ticket or a report, including ours. How the guard holds it.
What changed since 1.25.39
vet-skills diff mfa-servicenow-mcp 1.25.39 1.25.49In the sandbox, this version reached for nothing 1.25.39 did not. Compared from the two recorded runs, not from source.
1.25.39 was not run in the sandbox (install failed); the comparison covers what static analysis recorded only.
The guard can hold an update like this until someone approves it (runtime.on_change: hold). A change that only shows with a real credential, or on a path the run did not reach, is not in a trace and so not in this comparison.
Is [email protected] safe to install?
KillerSkills ran it in the KillerSkills honeypot sandbox (2 scenarios; not started: dependency install failed; the server itself was not observed) on 2026-11-19 and scored it 73/100 (Caution). It is not known to be malicious, but it has behaviour or code a security policy should decide on before install. Notable findings: Access to sensitive file paths; Access to sensitive file paths; Access to sensitive file paths. Summary: no network, touches nothing outside the project, writes no memory.
Behavior trace
Honeypot run · 1 s · 2 scenarios · 0 deviationsStatic findings
command = subprocess.run( # noqa: S603,S607 - fixed argv, no shellFix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.# from the real environment, not from .env).Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.# Do NOT close or kill it. Keep login_in_progress=True so thatFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.booting on Windows would kill the first host's in-flight login and then refuseFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell."""Ask a pid to exit. Never uses os.kill on Windows.Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.out = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.process = subprocess.Popen( # noqa: S603 - executable path comes from PlaywrightFix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out."review eval(), innerHTML, or Function() usage immediately."Fix: Put hosts, commands and strings in plain source. Decoding at runtime looks identical to malware and costs points on its own.subprocess.run( # noqa: S603,S607Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.out = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.crashed mid-login (kill -9, OOM, panic), our own `.lock` fileFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.# User-chosen base: create it private if missing, but NEVER chmod aFix: Request the narrowest permission that works; replace wildcards with the exact paths or hosts.(r"\beval\s*\(", "eval_usage", "Direct eval() — code injection risk"),Fix: Put hosts, commands and strings in plain source. Decoding at runtime looks identical to malware and costs points on its own.started = subprocess.run( # noqa: S603,S607 - fixed argv, no shellFix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.completed = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.# Load environment variables from .env fileFix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.subprocess.run( # noqa: S603,S607 - fixed argv, no shellFix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.# Attempt to load from .env if not provided via args/env vars directly in parse_argsFix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.not, and the user can always type the password themselves. Every outcome isFix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.subprocess.run(cmd, cwd=ROOT, env=env, check=True)Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.result = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.# first wait() — otherwise stop() would permanently kill keepalive.Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.subprocess.run(cmd, check=True, timeout=600)Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.# This used to add "CLAUDE.md asks authors to keep descriptions ≤80Fix: Remove the text that addresses the agent rather than the reader; describe behavior, do not instruct.# not permission to kill somebody's browser.Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.# Broad on purpose (one broken peer must never kill startup), butFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell."hint": "Script errors, warnings, gs.log output, platform messages",Fix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.logger.debug("Could not force-kill debug window pid %s: %s", pid, exc)Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.# var) must NOT kill server startup — the Playwright lesson: aFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.On Windows os.kill maps every signal other than CTRL_C/CTRL_BREAK to anFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.its own kill policy is how one of them ends up SIGKILLing a browser thatFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.Cost to the agent
Tokens estimated at four characters each. The tool list is re-read by the agent on every turn, so it is the number to compare between servers that do the same job: ranked across the index.
Known vulnerabilities
No advisories in OSV for mfa-servicenow-mcp@1.25.49 as of 16 h ago. This is the package itself, not its dependencies.
What would make this Verified
7 points to 80 · weights: behavior 45%, static 25%, prompt 15%, publisher 15%- Prompting user to enter sensitive credentials≈ +9 on the totalStop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
- Unclaimed publisher≈ +8 on the totalClaim jshsakura: a verified publisher gets +50 on publisher trust. It is free and usually instant with GitHub sign-in.
- Access to sensitive file paths≈ +4 on the totalStop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
- Python subprocess execution detected≈ +2 on the totalBuild commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
- Process kill command≈ +2 on the totalBuild commands from fixed arguments; do not interpolate untrusted strings into a shell.
- Python exec or eval of computed code≈ +2 on the totalPut hosts, commands and strings in plain source. Decoding at runtime looks identical to malware and costs points on its own.