Security audit

cpbl

Security checks across malware telemetry and agentic risk

Overview

The skill is mostly a CPBL sports lookup tool, but it explicitly teaches agents to use a stealth browser workflow to get around a third-party site's Anubis anti-bot protection.

Review this skill before installing if you are not comfortable with automated scraping or stealth-browser access to sites with anti-bot protections. For ordinary CPBL scores, schedules, standings, and stats, prefer the bundled official-site scripts; avoid using the twbsball StealthyFetcher workflow unless you have permission and understand the site's terms.

SkillSpector

By NVIDIA

Vulnerability Patterns

Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
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

Findings (6)

Lp3

Medium

Category: MCP Least Privilege
Confidence: 91% confidence
Finding: The skill clearly instructs use of local scripts, schedule cache files, and external network access, yet it declares no permissions. This creates a capability/permission mismatch that can bypass least-privilege controls and make users or the platform unaware that the skill can read local files, write cache data, and access external sites including stealth-fetching tools.

Context-Inappropriate Capability

Medium

Confidence: 92% confidence
Finding: The README explicitly promotes using Scrapling StealthyFetcher to bypass Anubis anti-bot and proof-of-work protections on a third-party site. Even though this is documentation rather than executable code, it equips the skill with operational guidance for evading access controls that are unrelated to the core CPBL lookup function and could be reused for unauthorized scraping.

Context-Inappropriate Capability

Medium

Confidence: 95% confidence
Finding: The installation steps and sample code provide a ready-made stealth browser workflow for fetching protected pages, including headless operation and waits for anti-bot challenges to complete. In the context of a sports information skill, this materially lowers the barrier to misuse by embedding evasion techniques that are unnecessary for normal end-user functionality.

Intent-Code Divergence

Medium

Confidence: 90% confidence
Finding: The document presents mutually inconsistent claims about whether CPBL standings data is actually retrievable and parsed successfully. In a skill that may be relied on by downstream agents or maintainers, this can cause unsafe operational assumptions, such as trusting empty or incomplete data as valid output, mis-prioritizing debugging, or deploying a skill believed to be functional when it is not.

Intent-Code Divergence

Medium

Confidence: 88% confidence
Finding: Claiming that all scripts function normally while later sections show incomplete implementations and API-blocked behavior is misleading documentation integrity failure. This is dangerous because users, reviewers, or automated systems may overtrust the skill's readiness, leading to incorrect outputs, failed automations, or concealment of real reliability limitations.

Description-Behavior Mismatch

Medium

Confidence: 94% confidence
Finding: The test log shows a mismatch between the skill’s advertised capabilities and its actual implementation: several commands return empty data or explicit 'not fully implemented' notes. This is a real security/reliability issue because an agent may route users to this skill expecting authoritative live scores, stats, or news, causing silent misinformation, failed automations, or unsafe downstream decisions based on incomplete results.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.