Back to skill

Security audit

Prompt Buttons

Security checks for vulnerabilities and agentic risk

Overview

This skill is a small button-prompt helper; the main issue is an extra hard-coded Chinese confirmation message, not a security threat.

Reasonable to install if you want Telegram-style prompt buttons. Review the hard-coded follow-up message and memory/subagent guidance if you need strict localization, quiet message behavior, or tight control over background work.

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
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • 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 (3)

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
98% confidence
Finding

The helper always sends the confirmation text in Chinese: '已发送按钮,请点数字选择。'. This imposes a specific language on all users with no configuration, negotiation, or documented locale constraint, which is a natural-language policy violation under the language/locale rule.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Low
Category
Not specified by scanner
Confidence
81% confidence
Finding

The file comment says the helper 'sends buttons and a short confirmation,' which implies a neutral acknowledgement. The actual code sends the specific message '已发送按钮,请点数字选择。' directing the user to choose by number, which is a stronger behavior than a generic confirmation and may be channel-inappropriate when inline buttons are available.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Low
Category
Not specified by scanner
Confidence
92% confidence
Finding

The manifest describes a helper that wraps prompts with tappable button menus or numbered-list fallbacks. This code not only sends the button prompt but also unconditionally sends a second confirmation message, which is additional user-facing behavior not implied by the stated purpose of merely wrapping prompts.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.