Back to skill

Security audit

loan-telemarketing-script

Security checks across malware telemetry and agentic risk

Overview

This skill should be reviewed carefully because its docs and code disagree, and it sends loan/customer inputs plus a user key to an insecure remote HTTP service.

Install only if you are comfortable with a third-party pay-per-use backend receiving the prompt data. Do not enter real customer PII, account numbers, contact lists, or confidential financial records. The publisher should fix the documentation/code mismatch, use HTTPS with a named trusted host, and clearly disclose exactly what data is sent and how the service key is used.

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • 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)

Lp3

Medium
Category
MCP Least Privilege
Confidence
92% confidence
Finding
The skill requires an environment variable for an external API key and explicitly instructs users to call a remote DeepSeek endpoint, but it declares no corresponding permissions or trust boundaries. This creates a transparency and governance gap: operators may not realize the skill can access secrets and transmit user-provided loan/customer data to a third-party service.

Description-Behavior Mismatch

Medium
Confidence
93% confidence
Finding
The skill advertises telemarketing-script generation, but the code also performs account registration, account info retrieval, billing-related messaging, and service catalog enumeration against a third-party backend. This expands the capability surface beyond the stated purpose and can mislead users into invoking remote commercial operations they did not expect, increasing privacy and trust risk.

Context-Inappropriate Capability

Medium
Confidence
88% confidence
Finding
The services command enumerates all AI services and pricing plans from the remote backend, including functionality unrelated to the declared loan telemarketing-script purpose. That creates an undisclosed discovery surface and may expose users to unrelated commercial endpoints or broaden the backend access available through the skill.

Missing User Warnings

Medium
Confidence
89% confidence
Finding
The skill is designed to generate telemarketing scripts from product information and customer profiles, which are likely to include sensitive financial or personal data, yet it lacks a clear warning to avoid submitting PII, account data, or regulated financial information. Because the workflow also sends prompts to an external model API, users may inadvertently disclose customer data to a third party, creating privacy, compliance, and data-handling risks.

Missing User Warnings

Medium
Confidence
96% confidence
Finding
The script sends user-supplied product, target-customer, and pain-point data plus the user account key to a remote API, but it does not provide an explicit consent or disclosure warning before transmission. In this skill context, those inputs may contain sensitive business, customer, or financial-marketing data, making undisclosed exfiltration to a third-party server more dangerous.

External Transmission

Medium
Category
Data Exfiltration
Content
if method == "GET":
            resp = requests.get(url, headers=h, params=data or {}, timeout=30)
        else:
            resp = requests.post(url, headers=h, json=data or {}, timeout=60)
        if resp.status_code >= 400:
            try:
                err = resp.json()
Confidence
97% confidence
Finding
The code performs external HTTP POST requests to a remote API and transmits user inputs and authentication material off-host. Because the default API base uses plain HTTP to a raw IP address, the transmission is vulnerable to interception or tampering in transit, which elevates this from ordinary external communication to a real security issue.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.