Skills/PyPI/mcp-sandbox-computer-vm-for-ai

mcp-sandbox-computer-vm-for-ai@0.2.3

VERIFIED· not started

Publisher claims: "Named Docker and Fly Machine sandbox computers with an MCP App dashboard" Observed in sandbox: no network, touches nothing outside the project, writes no memory.

Registry PyPI (MCP server)Publisher flujo-app · publishing for 103 daysInstalls 224Last detonated 6d ago
$claude mcp add mcp-sandbox-computer-vm-for-ai -- uvx mcp-sandbox-computer-vm-for-ai
WatchReport an issue

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: Verified 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
E2B_API_KEYFLY_API_TOKENFLY_TOKENKILNTAINERS_AUTH_TOKENMODAL_TOKEN_IDMODAL_TOKEN_SECRET or --e2b-api-key / --fly-token / --modal-token-id / --modal-token-secret / --auth-token / --allow-unauthenticated-httpConfiguration, not credentials: ENABLE_LIFECYCLE_TOOLS, FLY_APP_NAME, FLY_ORG, FLY_REGION, FLYCTL_INSTALL, KILNTAINERS_FLY_STATE_FILE, PROCESSOR_ARCHITECTURE, PROCESSOR_ARCHITEW6432
Where it talks to
no vendor host named in the codeAlso modal.com, fly.io, downloads.example — not the publisher's domain; check what for before connecting.
What the honeypot saw
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 modal.com, fly.io, downloads.example as well as its own vendor. The guard injects E2B_API_KEY, FLY_API_TOKEN, FLY_TOKEN, KILNTAINERS_AUTH_TOKEN, MODAL_TOKEN_ID, MODAL_TOKEN_SECRET only into requests to the host you bind (and its subdomains) and refuses the rest.

$vet-skills run --net declared -- uvx mcp-sandbox-computer-vm-for-ai

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 2026-11-08 and scored it 86/100 (Verified). In what ran, nothing we could detect went wrong: no undeclared network egress, no credential reads, no persistent instruction writes. What it does with a real credential was not observed; the static findings and the publisher carry the rest of the score. 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
0.2sprocessdependency install exit 1 ·
0.2sclaimednot started: dependency install failed (exit 1); the server itself was not observed
0.2sprocessscript backend runs setup only
0.2sprocessno install scripts to re-run

Static findings

MEDIUM
Python subprocess execution detectedsrc/kilntainers/backends/test_docker_integration.py:47 · Python subprocess execution detectedresult = subprocess.run([engine, "info"], capture_output=True, text=True)Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
MEDIUM
Process kill commandsrc/kilntainers/test_http_lifecycle.py:57 · Process kill command — agent can terminate processesos.kill = mock_kill # type: ignore[assignment]Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandsrc/kilntainers/test_lifecycle_integration.py:272 · Process kill command — agent can terminate processes# Actually kill the container externallyFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Python subprocess execution detectedsrc/kilntainers/test_cli_integration.py:22 · Python subprocess execution detectedresult = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
MEDIUM
Process kill commandsrc/kilntainers/test_lifecycle_integration.py:300 · Process kill command — agent can terminate processes"""Monitor death and kill externally."""Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandsrc/kilntainers/backends/test_e2b.py:677 · Process kill command — agent can terminate processes# kill should only be called onceFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Python subprocess execution detectedsrc/kilntainers/test_lifecycle_integration.py:102 · Python subprocess execution detectedresult = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
MEDIUM
Python subprocess execution detectedsrc/kilntainers/backends/test_docker_integration.py:213 · Python subprocess execution detectedsubprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
MEDIUM
Python subprocess execution detectedsrc/kilntainers/test_lifecycle_integration.py:125 · Python subprocess execution detectedresult = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
MEDIUM
Process kill commandsrc/kilntainers/test_http_lifecycle.py:50 · Process kill command — agent can terminate processes# Mock os.kill to capture callsFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandsrc/kilntainers/test_http_lifecycle.py:76 · Process kill command — agent can terminate processesos.kill = original_kill # type: ignore[assignment]Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Python subprocess execution detectedsrc/kilntainers/backends/test_docker_integration.py:263 · Python subprocess execution detectedresult = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
MEDIUM
Python subprocess execution detectedsrc/kilntainers/test_cli.py:680 · Python subprocess execution detectedresult = subprocess.run([sys.executable, "-c", script], check=False)Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
MEDIUM
Python subprocess execution detectedsrc/kilntainers/backends/test_docker_integration.py:225 · Python subprocess execution detectedsubprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
MEDIUM
Python subprocess execution detectedsrc/kilntainers/test_cli_integration.py:79 · Python subprocess execution detectedresult = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
MEDIUM
Python subprocess execution detectedsrc/kilntainers/test_lifecycle_integration.py:87 · Python subprocess execution detectedresult = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
MEDIUM
File type combined with upload actionsrc/kilntainers/test_http_lifecycle.py:302 · File type combined with upload action — agent may exfiltrate local data files# Client 2 should see b.txt in its ls outputFix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.
AST07
MEDIUM
Python subprocess execution detectedsrc/kilntainers/backends/test_docker_integration.py:252 · Python subprocess execution detectedresult = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
MEDIUM
Python subprocess execution detectedsrc/kilntainers/test_cli_integration.py:91 · Python subprocess execution detectedresult = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
MEDIUM
File type combined with upload actionsrc/kilntainers/test_http_lifecycle.py:292 · File type combined with upload action — agent may exfiltrate local data files# Client 1 should see a.txt in its ls outputFix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.
AST07
MEDIUM
Python subprocess execution detectedsrc/kilntainers/test_cli_integration.py:118 · Python subprocess execution detectedresult = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
MEDIUM
Python subprocess execution detectedsrc/kilntainers/test_lifecycle_integration.py:273 · Python subprocess execution detectedsubprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
MEDIUM
Process kill commandsrc/kilntainers/test_http_lifecycle.py:51 · Process kill command — agent can terminate processesoriginal_kill = os.killFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Python subprocess execution detectedsrc/kilntainers/test_lifecycle_integration.py:214 · Python subprocess execution detectedresult = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
MEDIUM
Python subprocess execution detectedsrc/kilntainers/test_cli_integration.py:103 · Python subprocess execution detectedresult = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
MEDIUM
Process kill commandsrc/kilntainers/test_lifecycle_integration.py:346 · Process kill command — agent can terminate processes# In HTTP mode, we can't easily kill the sandbox and detect deathFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05

Cost to the agent

Instructions loaded
≈63 tok
1 file · 56 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-sandbox-computer-vm-for-ai@0.2.3 as of 6d ago. This is the package itself, not its dependencies.