ClawHub
Back to skill

Security audit

AI Ball match analysis

Security checks for vulnerabilities and agentic risk

Overview

This skill is a read-only football match analysis helper that uses public AI Ball API data and explicitly avoids betting recommendations.

Before installing, understand that this skill can answer football match-probability questions using AI Ball's public record. It should be used as information only, not as betting advice, and users may want language preference handling tightened if they do not want English or Malay output by default.

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 (3)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The skill is scoped broadly enough to trigger on ordinary conversational requests like 'who will win X vs Y tonight,' which can cause the agent to engage in quasi-advisory sports outcome guidance without strong exclusion boundaries. Although the content repeatedly says not to recommend a side, the invocation surface is still wide and may lead the system to answer gambling-adjacent prompts in contexts where policy should require tighter gating.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The markdown directs the agent to always add lang=en so names come back in English, with Malay as an exception for Malay-speaking users. This is a natural-language locale policy constraint that forces a default language selection rather than offering the user a choice or deferring to user preference.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
88% confidence
Finding

The conventions table instructs consumers to use lang=en or lang=ms and says the default is Chinese but to 'always pass it.' This is a natural-language locale policy constraint that pushes a specific language selection without mentioning user preference, opt-in, or a justified region-specific requirement.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.