Security audit

Voice Call

Security checks for vulnerabilities and agentic risk

Overview

This voice-call skill does what it claims: it enables OpenClaw to place and manage phone calls through configured telephony providers, with expected privacy and consent cautions.

Install only if you want OpenClaw agents to be able to make or handle phone calls. Use mock mode for testing, require explicit confirmation before real outbound calls in your workflows, avoid speaking secrets unless necessary, and configure provider credentials, webhook exposure, inbound policy, session scope, and call-log storage deliberately.

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (1)

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 92% confidence
Finding: The skill explicitly enables placing real phone calls and sending spoken messages through third-party telephony providers, but the documentation does not warn that using the tool may contact external recipients or transmit message content and phone numbers to providers such as Twilio, Telnyx, or Plivo. This creates a real risk of unintended outreach, privacy violations, and accidental disclosure of sensitive data when an agent uses the skill without clear consent and data-handling guidance.
Content: No source excerpt is available for this finding.

Static analysis

Detected: suspicious.dangerous_exec

Shell command execution detected (child_process).

Critical

Code: suspicious.dangerous_exec
Location: dist/.setup/runtime-entry-CbSKxSKk.mjs:1890
Evidence: const proc = spawn("ngrok", args, {