ClawHub
Back to skill

Security audit

Alpha Feed Creator

Security checks across malware telemetry and agentic risk

Overview

This skill is not clearly malicious, but it asks agents to run scheduled collection, write shared reports, and push content to external groups without a precise input/output contract or strong user-control boundaries.

Install only if you intend to let an agent collect from external platforms and potentially send reports to team channels. Before enabling it, require dry-run output, explicitly review push targets, webhook tokens, schedule entries, and shared output paths, and avoid broad all-source scheduled runs until the configuration is narrowed.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Intent-Code Divergence

Medium
Confidence
93% confidence
Finding
The documented output schema describes a generic code-audit scoring report rather than the skill’s actual feed-collection, ranking, and push behavior. This mismatch can mislead users and downstream agents into trusting the wrong result structure, causing unsafe automation decisions, missed review of outbound pushes, or improper handling of collected content.

Intent-Code Divergence

Medium
Confidence
92% confidence
Finding
The input contract lists generic review fields like content and strict_level that do not match the skill’s advertised operational parameters such as sources, keywords, schedules, push targets, and output paths. In an agent setting, ambiguous or contradictory inputs can lead to unintended execution, malformed invocations, or user approval based on a false understanding of what the skill will do.

Missing User Warnings

Medium
Confidence
90% confidence
Finding
The skill promotes automatic group pushes, scheduled collection, and file writes, but does not present a clear upfront warning that it may perform recurring external actions and write data to shared locations. In practice this increases the risk of accidental data dissemination, spammy notifications, privacy issues, and persistent automated behavior being enabled without informed consent.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.