ClawHub
Back to skill

Security audit

AI行业新闻搜集小技能

Security checks for vulnerabilities and agentic risk

Overview

This skill is a simple AI-news summarization prompt that uses web fetching for current news and does not install code, persist data, or request privileged access.

Before installing, expect the skill to browse the web for AI news and return Chinese-formatted summaries. Users who want another language or tighter source/time constraints should state those explicitly when invoking it.

A.I.G

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

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 (2)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The skill can activate on a very broad class of requests about AI news without clear scope limits, exclusions, or disambiguation criteria. This can cause unintended invocation, over-collection via web fetching, and responses that bypass more appropriate skills or normal user clarification, increasing the risk of incorrect or undesired external data retrieval.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
84% confidence
Finding

The skill hard-codes Chinese-language behavior and output formatting without checking the user's preferred language. While not a direct security flaw, this can cause user-agent misalignment, reduce transparency, and make it harder for downstream reviewers or safety systems expecting another language to validate the output.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.