Skills/PyPI/mcp-ssh-live

mcp-ssh-live@0.1.0

CAUTION· not started

Publisher claims: "Interactive, streaming SSH tool for LLM agents via MCP (Model Context Protocol)" Observed in sandbox: no network, touches nothing outside the project, writes no memory.

Registry PyPI (MCP server)Publisher pmboxbiz · publishing for 196 daysInstalls 18Last detonated 5d ago
What would fix thisWatchDispute verdict

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
no credential variables read in the code
Where it talks to
no vendor host named in the code
What the honeypot saw
no credential files read · no network
What we could not test
server startup · the tool list · tool behavior

Connect it under the guard. Not started: dependency install failed; the server itself was not observed. The code names no credential variable; the guard pins the network to what the honeypot saw and logs every call.

$vet-skills run --net declared -- uvx mcp-ssh-live

Never paste a credential into a chat, a ticket or a report, including ours. How the guard holds it.

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 2026-11-09 and scored it 68/100 (Caution). It is not known to be malicious, but it has behaviour or code a security policy should decide on before install. Notable findings: Destructive file system command detected; References a credential file in the home directory; Privileged sudo command. Summary: no network, touches nothing outside the project, writes no memory.

Behavior trace

Honeypot run · 0 s · 2 scenarios · 0 deviations
t+CategoryObserved action
0.0sprocessno install scripts in bundle
0.1sprocessdependency install exit 1 ·
0.1sclaimednot started: dependency install failed (exit 1); the server itself was not observed
0.1sprocessscript backend runs setup only
0.1sprocessno install scripts to re-run

Static findings

MEDIUM
File permission modification commandsrc/mcp_ssh_live/tools/sftp_tool.py:391 · File permission modification command — agent changes file ownership/permissions# chmod failure is non-fatal — the bytes are alreadyFix: Request the narrowest permission that works; replace wildcards with the exact paths or hosts.
AST05
HIGH
Destructive file system command detectedsrc/mcp_ssh_live/tools/persistent_tool.py:216 · Destructive file system command detected" ssh_exec('rm -rf /tmp/mcp-persistent/<id>')"Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandsrc/mcp_ssh_live/connection.py:525 · Process kill command — agent can terminate processes``kill -TERM/-KILL`` on a separate short session. This sidesteps theFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandsrc/mcp_ssh_live/job_manager.py:324 · Process kill command — agent can terminate processes# the dependency would be circular, and the kill command weFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandsrc/mcp_ssh_live/job.py:487 · Process kill command — agent can terminate processes``ssh_remove_job`` and during server shutdown. Does NOT kill theFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandsrc/mcp_ssh_live/connection.py:621 · Process kill command — agent can terminate processes# Try to kill the remote process. We need a separate session;Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
HIGH
References a credential file in the home directorysrc/mcp_ssh_live/tools/connect_tool.py:175 · References a credential file in the home directory"default key files (~/.ssh/id_*).\n\n"Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
AST03
MEDIUM
File permission modification commandsrc/mcp_ssh_live/tools/sftp_tool.py:396 · File permission modification command — agent changes file ownership/permissions"ssh_upload: chmod %o on %r failed: %s",Fix: Request the narrowest permission that works; replace wildcards with the exact paths or hosts.
AST05
HIGH
Privileged sudo commandsrc/mcp_ssh_live/tools/stdin_tool.py:303 · Privileged sudo command — agent may execute with elevated privileges"program sees a complete line (this is what 'sudo -S' and "Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandsrc/mcp_ssh_live/job_manager.py:126 · Process kill command — agent can terminate processesvia the standard Phase 3 kill path — wired up by the signal tool).Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandsrc/mcp_ssh_live/connection.py:727 · Process kill command — agent can terminate processes"""Send ``kill -SIGNAL pid`` on a fresh short session.Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
HIGH
Privileged sudo commandsrc/mcp_ssh_live/tools/spawn_tool.py:182 · Privileged sudo command — agent may execute with elevated privilegesprograms (``sudo -S``, REPLs, ``passwd``) and forFix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
MEDIUM
Process kill commandsrc/mcp_ssh_live/tools/jobs_tool.py:34 · Process kill command — agent can terminate processesdies — for a hard kill the LLM should call ``ssh_signal(KILL)`` firstFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandsrc/mcp_ssh_live/connection.py:536 · Process kill command — agent can terminate processescapture_pid: wrap the command so we can kill it by PID on timeout.Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandsrc/mcp_ssh_live/job_manager.py:393 · Process kill command — agent can terminate processes"""Send ``kill -<SIG> <pid>`` to the remote on a fresh exec channel.Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandsrc/mcp_ssh_live/tools/jobs_tool.py:297 · Process kill command — agent can terminate processes"NOT guarantee the remote process dies — for a hard kill "Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandsrc/mcp_ssh_live/server.py:150 · Process kill command — agent can terminate processes# send ``kill -TERM``/``-KILL`` to each live job's remote pid viaFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
File permission modification commandsrc/mcp_ssh_live/tools/sftp_tool.py:555 · File permission modification command — agent changes file ownership/permissions"the remote after the transfer. A chmod failure is "Fix: Request the narrowest permission that works; replace wildcards with the exact paths or hosts.
AST05
HIGH
Access to sensitive file pathssrc/mcp_ssh_live/tools/connect_tool.py:175 · Access to sensitive file paths — SSH keys, AWS credentials, env files, or system files"default key files (~/.ssh/id_*).\n\n"Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
AST05
MEDIUM
File permission modification commandsrc/mcp_ssh_live/tools/sftp_tool.py:557 · File permission modification command — agent changes file ownership/permissions"in the response means the chmod was skipped or rejected."Fix: Request the narrowest permission that works; replace wildcards with the exact paths or hosts.
AST05
MEDIUM
Process kill commandsrc/mcp_ssh_live/connection.py:524 · Process kill command — agent can terminate processes``capture_pid=False``), so we can kill it on timeout viaFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandsrc/mcp_ssh_live/job.py:455 · Process kill command — agent can terminate processesDoes NOT kill the remote process — that's ``ssh_signal``'s job.Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandsrc/mcp_ssh_live/job_manager.py:294 · Process kill command — agent can terminate processesa short ``kill -TERM <pid>`` exec on the same transport — sameFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandsrc/mcp_ssh_live/job.py:771 · Process kill command — agent can terminate processes# a kill path that already set status=KILLED isn'tFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandsrc/mcp_ssh_live/tools/jobs_tool.py:117 · Process kill command — agent can terminate processesprocess alive, and the LLM should either accept that or kill it.Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
HIGH
Access to sensitive file pathssrc/mcp_ssh_live/__main__.py:18 · Access to sensitive file paths — SSH keys, AWS credentials, env files, or system files--host stage=10.0.0.5 --user deploy --key ~/.ssh/stage_keyFix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
AST05
HIGH
Privileged sudo commandsrc/mcp_ssh_live/tools/stdin_tool.py:4 · Privileged sudo command — agent may execute with elevated privileges§"Edge cases & decisions" → "2. Interactive input (sudo password)" forFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
File permission modification commandsrc/mcp_ssh_live/tools/sftp_tool.py:554 · File permission modification command — agent changes file ownership/permissions"Optional mode (octal, e.g. 493 for 0o755) runs chmod on "Fix: Request the narrowest permission that works; replace wildcards with the exact paths or hosts.
AST05
HIGH
Privileged sudo commandsrc/mcp_ssh_live/tools/stdin_tool.py:19 · Privileged sudo command — agent may execute with elevated privilegesto pre-bake the input into the command (``echo pw | sudo -S …``),Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
HIGH
Privileged sudo commandsrc/mcp_ssh_live/tools/stdin_tool.py:12 · Privileged sudo command — agent may execute with elevated privileges* ``sudo -S -p ''`` reads the password from stdin (one line).Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandsrc/mcp_ssh_live/job.py:33 · Process kill command — agent can terminate processes(Phase 3) for the ``kill -SIGNAL <pid>`` fallback on OpenSSH serversFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
HIGH
Privileged sudo commandsrc/mcp_ssh_live/tools/stdin_tool.py:299 · Privileged sudo command — agent may execute with elevated privileges"by ssh_spawn). Used for password prompts (sudo -S -p ''), "Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandsrc/mcp_ssh_live/job_manager.py:390 · Process kill command — agent can terminate processes# ------------------------------------------------------------------ graceful kill helperFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandsrc/mcp_ssh_live/tools/spawn_tool.py:257 · Process kill command — agent can terminate processes# (used by ssh_signal to fall back to ``kill -SIGNAL <pid>``Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
MEDIUM
Process kill commandsrc/mcp_ssh_live/tools/signal_tool.py:41 · Process kill command — agent can terminate processeswhich includes the spawned job. We wait up to ~500 ms for the killFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandsrc/mcp_ssh_live/tools/signal_tool.py:14 · Process kill command — agent can terminate processespresent on real servers. Relying on ``send_signal`` alone means "kill myFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandsrc/mcp_ssh_live/tools/signal_tool.py:38 · Process kill command — agent can terminate processes2. **``kill -<SIG> <pid>``** fallback — opened as a brand-new shortFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandsrc/mcp_ssh_live/tools/signal_tool.py:113 · Process kill command — agent can terminate processes# shells accept both ``kill -TERM`` and ``kill -15``, so we normalizeFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandsrc/mcp_ssh_live/tools/signal_tool.py:111 · Process kill command — agent can terminate processes# * The wire format on the remote is ``kill -<NAME>`` where ``<NAME>``Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05

Cost to the agent

Instructions loaded
≈56 tok
1 file · 26 in bundle
Tool list per prompt
≈0 tok
0 tools · 0 per tool
Handshake
no reply
launch to tools/list, stdio
Undocumented tools
0
description under 20 chars · avg 0

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 mcp-ssh-live@0.1.0 as of 5d ago. This is the package itself, not its dependencies.

What would make this Verified

12 points to 80 · weights: behavior 45%, static 25%, prompt 15%, publisher 15%
  1. Unclaimed publisher≈ +8 on the totalClaim pmboxbiz: a verified publisher gets +50 on publisher trust. It is free and usually instant with GitHub sign-in.
  2. Destructive file system command detected≈ +4 on the totalBuild commands from fixed arguments; do not interpolate untrusted strings into a shell.
  3. References a credential file in the home directory≈ +4 on the totalStop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
  4. Privileged sudo command≈ +4 on the totalBuild commands from fixed arguments; do not interpolate untrusted strings into a shell.
  5. 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.
  6. File permission modification command≈ +2 on the totalRequest the narrowest permission that works; replace wildcards with the exact paths or hosts.
Publish a new version to PyPI, then queue it here. Not your skill? Claim the handle to get verdict changes by email.Claim pmboxbizPropose a patch to the maintainer