Security audit

品管圈技能

Security checks for vulnerabilities and agentic risk

Overview

The skill is a coherent QCC presentation assistant that asks users for project materials, with one usability/privacy note about broad trigger phrases.

Before installing, be aware that the skill may activate for generic presentation or speech-writing requests; use it when you intend to work with QCC project materials, and avoid providing unrelated private documents if the conversation is not about a QCC project.

SkillSpector

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 (1)

Vague Triggers

Medium

Confidence: 92% confidence
Finding: The trigger phrases include broad everyday expressions such as '写讲稿' and '发表前检查', which can cause the skill to auto-load in contexts beyond QCC project publishing. Unintended activation increases the chance of the agent requesting files, paths, or project artifacts in situations where this skill is not appropriate, leading to workflow confusion and possible over-collection of user data.

Static analysis

No suspicious patterns detected.