Back to skill

Security audit

home-router-security-auditor

Security checks for vulnerabilities and agentic risk

Overview

This skill provides purpose-aligned router guidance and network inventory, but device identification is manual and some recommended changes require recovery precautions.

Use inventory only on networks you control, and identify devices manually; the default inventory uses cached entries, while the optional sweep covers a limited range. Verify settings yourself before treating the baseline as accurate. Establish working wired access before disabling Wi-Fi administration. Before resetting or reflashing, obtain model-specific recovery instructions and necessary ISP settings, and plan for connectivity loss. Keep passwords out of baseline notes. These operational limitations do not indicate malicious behavior.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • 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 (10)

Skill declares no tool scope ('permissions' or 'allowed-tools') but code capabilities were detected: file_write, network, shell.

Medium
Category
MCP Least Privilege
Confidence
70% confidence
Finding

Without declared permissions the skill's intent is opaque and cannot be validated.

Content

No source excerpt is available for this finding.

The instruction to restrict router administration to wired clients omits a warning about losing administrative access over Wi-Fi.

Medium
Category
Output Handling
Confidence
94% confidence
Finding

Restricting router administration to wired clients can lock out a user who is performing the hardening over Wi-Fi and lacks a working Ethernet connection. The restriction is a legitimate security measure, but omitting a connectivity prerequisite creates a conditional administrative-availability risk, potentially requiring physical recovery or a factory reset.

Content

No source excerpt is available for this finding.

The compromise-response workflow instructs a factory reset without warning about configuration loss or loss of connectivity.

Medium
Category
Output Handling
Confidence
88% confidence
Finding

Factory resetting or reflashing a router can erase ISP credentials, VLAN settings, and other configuration needed to restore connectivity, potentially leaving users offline or disconnecting dependent devices. The compromise-response context makes these actions legitimate, but this civilian-oriented guide omits warnings and recovery prerequisites that inexperienced users need. This is an operational-safety omission rather than evidence of malicious intent.

Content

No source excerpt is available for this finding.

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/router_audit.py (reported line 145)May include surrounding context.

python
pass
    # 3. platform commands
    if platform.system() == "Windows":
        out = subprocess.run(["ipconfig"], capture_output=True, text=True).stdout
        m = re.search(r"IPv4[^\d]+(\d+\.\d+\.\d+)\.\d+", out)
        return m.group(1) if m else "192.168.1"
    for cmd in (["ip", "-4", "route", "get", "8.8.8.8"], ["ifconfig"]):

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/router_audit.py (reported line 150)May include surrounding context.

python
return m.group(1) if m else "192.168.1"
    for cmd in (["ip", "-4", "route", "get", "8.8.8.8"], ["ifconfig"]):
        try:
            out = subprocess.run(cmd, capture_output=True, text=True).stdout or ""
            m = re.search(r"src (\d+\.\d+\.\d+\.\d+)", out)
            if m:
                return m.group(1).rsplit(".", 1)[0]

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/router_audit.py (reported line 175)May include surrounding context.

python
if rows:
            return rows
    if platform.system() == "Windows":
        out = subprocess.run(["arp", "-a"], capture_output=True, text=True).stdout
        for line in out.splitlines():
            m = re.match(r"\s*(\d+\.\d+\.\d+\.\d+)\s+([0-9a-fA-F-]{17})", line)
            if m:

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/router_audit.py (reported line 182)May include surrounding context.

python
rows.append((m.group(1), m.group(2).replace("-", ":").lower()))
    else:
        try:
            out = subprocess.run(["arp", "-an"], capture_output=True, text=True).stdout
            for line in out.splitlines():
                m = re.search(r"\((\d+\.\d+\.\d+\.\d+)\) at ([0-9a-f:]{17})", line)
                if m:

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/router_audit.py (reported line 197)May include surrounding context.

python
to = "-w" if platform.system() == "Windows" else "-W"
    to_ms = str(int(timeout_s * 1000)) if platform.system() == "Windows" else str(int(timeout_s))
    try:
        r = subprocess.run(
            ["ping", flag, str(count), to, to_ms, host],
            capture_output=True, text=True, timeout=timeout_s + 2,
        )

subprocess module call

Medium
Category
Dangerous Code Execution
Confidence
70% confidence
Finding

subprocess module calls execute external commands. Without careful input validation, this enables command injection.

Content

Scanner excerpt · scripts/router_audit.py (reported line 208)May include surrounding context.

python
def resolve_name(ip):
    try:
        out = subprocess.run(
            ["ping", "-a" if platform.system() == "Windows" else "-c", "1",
             ip] if platform.system() == "Windows" else ["getent", "hosts", ip],
            capture_output=True, text=True, timeout=2,

The inventory lists devices but does not flag unknown clients as promised by the manifest.

Low
Category
Excessive Agency
Confidence
99% confidence
Finding

The manifest explicitly promises a local-network inventory with flagging of unknown clients. At L244–L250, the code prints IP addresses, MAC addresses, and names without comparing clients against known devices or assigning any status; L254–L260 merely prints a generic verdict legend for manual interpretation.

Content

No source excerpt is available for this finding.