ClawHub
Back to skill

Security audit

Red Alert (Israel)

Security checks across malware telemetry and agentic risk

Overview

This alert skill has a legitimate emergency-notification purpose, but its main skill description under-discloses Docker, Home Assistant, 3CX calling, credential storage, and background persistence.

Review carefully before installing. Only use this if you intend to run a persistent alert monitor with Docker, OpenClaw WhatsApp posting, optional 3CX calls, and optional Home Assistant TTS. Configure all destinations explicitly, remove hard-coded defaults, restrict .env permissions, inspect the crontab entry, and avoid providing a Home Assistant token unless you trust the host and need speaker announcements.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Taint TrackingDirect Taint Flow, Variable-Mediated Taint Flow, Credential Exfiltration Chain
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (6)

Tainted flow: 'cmd' from os.getenv (line 92, credential/environment) → subprocess.run (code execution)

Medium
Category
Data Flow
Content
"--channel", "whatsapp",
               "--target", target,
               "--message", message]
        r = subprocess.run(cmd, capture_output=True, text=True, timeout=15)
        if r.returncode == 0:
            log.info(f"✅ WhatsApp → {target}")
        else:
Confidence
80% confidence
Finding
r = subprocess.run(cmd, capture_output=True, text=True, timeout=15)

Tainted flow: 'HA_URL' from os.getenv (line 30, credential/environment) → requests.post (network output)

Critical
Category
Data Flow
Content
def ha_tts(text: str):
    """הכרז קולית דרך רמקול Home Assistant"""
    try:
        r = requests.post(
            f"{HA_URL}/api/services/tts/google_translate_say",
            headers={"Authorization": f"Bearer {HA_TOKEN}", "Content-Type": "application/json"},
            json={"entity_id": HA_TTS_SPEAKER, "message": text, "language": "iw"},
Confidence
87% confidence
Finding
r = requests.post( f"{HA_URL}/api/services/tts/google_translate_say", headers={"Authorization": f"Bearer {HA_TOKEN}", "Content-Type": "application/json"}, j

Tainted flow: 'OREF_API' from os.getenv (line 24, credential/environment) → requests.get (network output)

Critical
Category
Data Flow
Content
def check_alert():
    global last_alert_id, alert_sent_at, all_clear_sent
    try:
        data = requests.get(OREF_API, timeout=5).json()
    except Exception as e:
        log.warning(f"⚠️ API: {e}")
        return
Confidence
82% confidence
Finding
data = requests.get(OREF_API, timeout=5).json()

Lp3

Medium
Category
MCP Least Privilege
Confidence
93% confidence
Finding
The skill documentation indicates capabilities to access environment variables, make network requests, and invoke shell commands, but it declares no permissions. This creates a transparency and trust problem: operators cannot accurately assess the skill's effective privileges, and the hidden shell/network capability materially increases the blast radius if the skill is modified, compromised, or misconfigured.

Tp4

High
Category
MCP Tool Poisoning
Confidence
97% confidence
Finding
The documented purpose claims an OpenClaw-only design with no Home Assistant, wacli, or Docker monitor, yet the analyzed behavior includes additional undeclared components such as Docker, Home Assistant, and 3CX calling. This mismatch is dangerous because it conceals actual attack surface, secrets handling, and external integrations from reviewers, making unsafe deployment more likely and enabling privileged functionality that users did not consent to.

Missing User Warnings

Medium
Confidence
98% confidence
Finding
The installer prompts for a Home Assistant access token and writes it in plaintext to a local .env file without warning, permission hardening, or secure secret storage. On multi-user systems or systems with weak file permissions, this can expose a powerful automation credential that may allow an attacker to read state, control devices, or pivot into the home automation environment.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.