Back to skill

Security audit

requesting-code-revi

Security checks for vulnerabilities and agentic risk

Overview

This skill has no executable payload, but its code-review purpose is broadened into generic development automation with command execution and vague API credential use.

Install only if you are comfortable with a broadly scoped development automation skill that may read project files and run commands. For safer use, keep it limited to explicit code-review tasks, avoid providing secrets or API keys unless necessary, and review any command before execution.

Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Vague Triggers

High
Category
Not specified by scanner
Confidence
92% confidence
Finding

The activation/description text is overly broad, covering code generation, debugging, deployment, automation, data analysis, and workflow orchestration rather than a narrowly scoped code-review function. This can cause the skill to be invoked in unintended contexts and gain access to sensitive development inputs or influence actions outside its stated purpose, especially since the skill advertises read and exec capabilities.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The scenario section expands the skill into general development, operations, and content-creation use cases, which materially exceeds a code-review skill's stated role. In practice this broadens where the agent may apply the skill, increasing the chance of misuse, unnecessary data exposure, or inappropriate execution in contexts that may involve privileged files or commands.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
80% confidence
Finding

The description states that the skill supports Chinese interaction without documenting language choice or fallback behavior, which can override user expectations and reduce transparency. While not a direct code-execution flaw, forced or implicit language constraints can degrade review quality, cause misunderstandings in security-sensitive output, and lead users to approve or act on misunderstood results.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
79% confidence
Finding

The core feature statement again implies Chinese interaction as a built-in default without an opt-in mechanism or user-facing control. In a code-review context, ambiguity about language can affect comprehension of findings, remediation, or warnings, making the skill less reliable and potentially unsafe in multilingual teams.

Content

No source excerpt is available for this finding.