ClawHub
Back to skill

Security audit

Daily Ai News Skill Litiao

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed AI news briefing helper that searches and summarizes public web sources without persistence or destructive behavior.

Before installing, expect this skill to send AI-news search queries and article URLs to external search or web-reading services, and only use it in environments where outbound browsing and Tavily API key use are acceptable.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Missing User Warnings

Medium
Confidence
95% confidence
Finding
The description does not warn that the skill will perform web searches and retrieve content from third-party sites, including use of external search providers. Without clear disclosure, users may unknowingly trigger outbound requests, which is a transparency and privacy issue and can be risky in environments where browsing should be explicit.

Vague Triggers

Medium
Confidence
93% confidence
Finding
The activation conditions are broad enough to match ordinary AI-related conversation, which can cause the skill to trigger and perform external searches or fetch third-party content when the user did not clearly request browsing. In an agent environment, overbroad triggers increase the chance of unintended tool use, unnecessary data sharing to external services, and confusion about why networked actions occurred.

Natural-Language Policy Violations

Low
Confidence
68% confidence
Finding
The file includes Chinese trigger phrases as explicit invocation examples, but it does not clarify whether responses will follow the user's language or whether multilingual behavior is supported generally. This could create ambiguity around language handling and locale expectations.

Context-Inappropriate Capability

Low
Confidence
87% confidence
Finding
The manifest describes a skill for aggregating and summarizing latest AI news from news websites and web search, with concise briefs and direct article links. At L280-L281, the documented follow-up behavior adds searching for Twitter reactions and analysis pieces, which is a different capability centered on opinion gathering rather than straightforward news briefing.

Static analysis

No suspicious patterns detected.