Back to skill

Security audit

Xueersi Daily Question Generator

Security checks for vulnerabilities and agentic risk

Overview

This is a simple study-question skill with no code execution, external access, credential handling, or persistence beyond conversational progress tracking.

Installers should be aware that generic requests like "give me a question" may route to this skill; use a more specific invocation if multiple education skills are installed. No sensitive access or high-impact authority was found.

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)

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
83% confidence
Finding

The description limits the skill to Chinese, Math, and English and presents bilingual branding, but does not state whether the interaction language is user-selectable or why a fixed locale/language scope is required. Under the policy, language or locale constraints should be optional for the user or clearly justified.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The trigger phrases "daily question" and especially "give me a question" are broad and likely to match ordinary conversation outside the intended educational workflow. This can cause unintended skill activation, hijacking user intent or suppressing other more appropriate skills, which is a genuine routing and UX security concern even though the skill content itself is low risk.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.