Back to skill

Security audit

AI Call Opt-Out and Pre-Call Check

Security checks for vulnerabilities and agentic risk

Overview

ClawScan rates this skill suspicious with medium confidence because its examples send phone and call outcome data to an external service without a privacy notice or sharing choice, and one integration may still call opted-out numbers.

Before installing, consider whether you approve sending phone numbers and call-related details to the named external service, and whether those recipients have consented to that sharing. Check that any integration enforces opt-outs before dialing. The scan is partial because two references could not be resolved, so the evidence is incomplete.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (8)

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 67)May include surrounding context.

bash
# 1. before the call
curl "https://linepref.byronwebstudio.com.au/v1/check?number=%2B15125550100&website=example.com"

# 3. key press or spoken opt-out
curl -X POST https://linepref.byronwebstudio.com.au/v1/optout -H 'content-type: application/json' \

The workflow sends called parties' phone numbers and call-related data to a third-party service without a privacy warning.

Medium
Category
Output Handling
Confidence
95% confidence
Finding

As a standalone recipe, this source sends destination numbers to an external provider and subsequently sends opt-out and outcome information without a privacy notice or authorization prerequisite. The legitimate suppression-list purpose reduces suspicion of malicious intent, but does not eliminate the risk of disclosing contact and calling-activity data outside the user's approved systems.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · references/bland.md (reported line 9)May include surrounding context.

do_not_call:

bash
curl "https://linepref.byronwebstudio.com.au/v1/check?number=%2B15125550100&website=example.com"

During the call (webhook tool)

The integration sends phone numbers and call metadata to an external service without a privacy disclosure.

Medium
Category
Output Handling
Confidence
94% confidence
Finding

The standalone integration passes the called number and optional website to the service and unconditionally reports the number with call timing and outcome metadata. It provides no privacy disclosure or reporting opt-in, so adopting the example can expose contact and operational data without an informed sharing decision.

Content

No source excerpt is available for this finding.

The workflow sends call-related data to an external MCP service without a privacy warning.

Medium
Category
Output Handling
Confidence
94% confidence
Finding

The proposed system prompt mandates external checks and outcome reporting for every call without first establishing approval for sharing call-related data. Although the MCP endpoint is visible, identifying a server is not equivalent to explaining its data collection or offering a reporting choice.

Content

No source excerpt is available for this finding.

The integration provides phone numbers and call outcomes to Linepref without a privacy disclosure.

Medium
Category
Output Handling
Confidence
86% confidence
Finding

The standalone example sends phone numbers and optional websites through the client and reports call outcomes without a privacy notice or approval check. Using a client library makes the external transfer less apparent than a direct URL, while the unconditional reporting can disclose calling activity that the user did not agree to share.

Content

No source excerpt is available for this finding.

The dashboard-only integration calls opted-out numbers instead of skipping the call.

Medium
Category
Excessive Agency
Confidence
99% confidence
Finding

The dashboard-only alternative initiates the call before acting on a do_not_call result, so an opted-out recipient can still be contacted. Ending the call immediately or apologizing does not enforce pre-dial suppression and can cause unwanted calls and compliance exposure.

Content

No source excerpt is available for this finding.

The workflow sends phone numbers and call outcomes to an external service without a privacy warning.

Medium
Category
Output Handling
Confidence
98% confidence
Finding

The post-call webhook sends the recipient's number and categorized call outcome to an external provider without a privacy disclosure or a reporting choice in this source. These linked fields reveal contact activity and potentially business interactions, even though no recording or transcript is transmitted.

Content

No source excerpt is available for this finding.