Back to skill

Security audit

复习规划师

Security checks for vulnerabilities and agentic risk

Overview

ClawScan rates this skill benign with medium confidence: scanners found no AIG issues and one scope concern about generating example questions beyond its stated planning role.

The verdict is benign, not a guarantee of no risk. Before installing, consider whether you are comfortable with the skill potentially generating individual example questions despite its stated planning boundary. Confidence is medium, and Skillspector's analysis was partial because three local references could not be resolved.

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 (1)

The skill permits generating concrete example and variant questions despite its manifest limiting output to review plans and delegating concrete questions to other skills.

Medium
Category
Excessive Agency
Confidence
96% confidence
Finding

The manifest at L003 explicitly says this skill produces plans rather than assignments and routes concrete questions to assignment-design skills. However, L046–L047 instruct this skill to generate example or variant questions within review activities, requiring only a self-check and an AI-generated label; delegation is narrowed to complete question sets. This creates an actionable exception to the manifest's stated output boundary.

Content

No source excerpt is available for this finding.