Back to skill

Security audit

qa-quality-metrics

Security checks for vulnerabilities and agentic risk

Overview

This is a QA metrics guidance skill with no automatic execution or hidden data handling, though its optional unpinned install suggestion deserves caution.

Installers should treat this as benign QA guidance, but should not run the optional npx companion-install command unless they trust the Kokxi/qa-test-skills source and preferably pin or verify the exact version.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (6)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

L04 lists broad triggers such as “质量数据”, “趋势分析”, and “向管理层展示质量数据时”, alongside a general 'use when' description. These phrases are common in normal project conversations and the file does not provide exclusion conditions or negative examples to clarify when the skill should not activate.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

L10 says the skill should be used when the user mentions phrases like “趋势分析” or when they 'need to build a metrics system' or present quality data to management. This combines specific and very general conditions, but does not define what kinds of trend analysis or management reporting are in scope versus out of scope.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The manifest output-format explicitly says coverage conclusions must label scope and prohibits absolute expressions like '全覆盖/100%'. However, the report template later presents '需求覆盖率:[X]% (目标100%)', which directly encourages the kind of absolute coverage target the documentation says to avoid.

Content

No source excerpt is available for this finding.

Rp1

Medium
Category
MCP Rug Pull
Confidence
88% confidence
Finding

The skill instructs users to install an external package via npx skills add Kokxi/qa-test-skills without pinning a version or commit. That creates a supply-chain risk: a future package update or compromise could change what gets fetched and executed, and npx commonly runs retrieved code during installation. In this context the risk is real, though it appears operationally convenient rather than malicious.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
61% confidence
Finding

L04 presents the primary description in Chinese with an English sentence appended, but nowhere indicates that the user may choose their preferred language. This can violate a language/locale choice policy when a skill appears to assume a default language rather than explicitly honoring user preference.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
89% confidence
Finding

This markdown file presents user-facing instructional content entirely in Chinese and does not indicate that the language choice is optional or region-specific. Under the policy rule for language or locale constraints, forcing a single language without user opt-in can be a natural-language policy violation.

Content

No source excerpt is available for this finding.