Back to skill

Security audit

PowPow สมัครสมาชิก — สมัครฟอง ปักบันทึกการเดินทางลงแผนที่ คุยกับมนุษย์ดิจิทัล

Security checks for vulnerabilities and agentic risk

Overview

This is a guidance-only Thai-language PowPow registration helper with public product links and no local execution, credential collection, persistence, or hidden authority.

Installers should understand this skill is mainly for Thai-language PowPow onboarding and will direct users to PowPow registration/subscription pages plus public screenshot links. Users should complete registration and payment only on the official PowPow pages and should not share passwords, verification codes, or payment proof in chat.

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
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (6)

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The skill description and invocation examples are written as if the skill will operate in Thai by default, while also serving a global website. There is no statement that users may choose another language or that Thai-only guidance is an intentional, justified regional limitation, which creates a natural-language locale policy concern.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The manifest explicitly sets "language": "th", which indicates a fixed language constraint. The file does not offer any user opt-in or explain a region-specific reason for requiring Thai, so this appears to violate the language/locale policy criteria.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The heading 'บทสรุป (พูดคำตอบก่อน)' and the document’s instructions are entirely in Thai, indicating the skill is intended to respond in Thai by default. There is no indication that users may choose another language or that the Thai-only constraint is required for a region-specific compliance reason.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

This markdown file contains explicit user-facing text in Chinese, and the error table continues to present original Chinese messages as the canonical output. Under the policy, forcing a specific language without user opt-in can be a natural-language policy violation unless the locale restriction is clearly justified, which is not stated here.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

All instructional content in the file is presented in Thai, which can effectively force a specific language for operators or downstream use. The document does not indicate that Thai is optional, user-selected, or required for a documented region-specific purpose.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The guidance says to send direct screenshot links whenever the user 'doesn't know bubbles' or 'needs to see the product,' which is a vague trigger that can cause the agent to disclose external content too readily. Broad screenshot-invocation rules increase the chance of unnecessary data sharing, off-topic media injection, and inconsistent behavior, especially because the links point to third-party hosted resources outside the skill bundle.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.