Back to skill

Security audit

safety-net

Security checks for vulnerabilities and agentic risk

Overview

This is a Markdown-only safety checklist with some unclear trigger wording, but no executable code, hidden access, persistence mechanism, or data handling.

Install only if Chinese-language safety guidance is suitable for your workflow. Treat the daily automatic review language as aspirational unless the hosting platform provides that trigger, and rely on explicit user-controlled invocation for high-risk actions.

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 (2)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The skill defines contradictory activation semantics: it says the system will auto-trigger a daily compliance review, but elsewhere says the skill is only invoked on demand before high-risk actions. In a safety gate, ambiguity about when checks run can cause risky operations to bypass review entirely or create a false sense of protection, especially because the skill claims authority over safety rules and risk judgments.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
81% confidence
Finding

SQP-3 适用于所有文件。该技能说明全文均以中文撰写,未向用户提供语言偏好选择,也未说明这是特定地区或特定受众限定的技能;标题虽含英文名,但实际操作说明仅支持中文,可能构成未经用户选择的语言限定。

Content

No source excerpt is available for this finding.