ClawHub
Back to skill

Security audit

Wadiz DM Pipeline

Security checks across malware telemetry and agentic risk

Overview

This skill is a user-run crawler that drafts outreach messages, with some disclosure gaps but no hidden sending, persistence, destructive action, or private-data harvesting evident.

Install only if you are comfortable running a local browser crawler, installing Playwright/Chromium and anthropic dependencies, and potentially sending scraped public project text to Anthropic when an API key is set. Treat the CSV as draft outreach only: review every message, respect Wadiz/Tumblbug terms and anti-spam rules, and unset ANTHROPIC_API_KEY if you want template-only generation.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • 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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (4)

Lp3

Medium
Category
MCP Least Privilege
Confidence
89% confidence
Finding
The skill advertises executable behavior that can access environment data and write files, but it does not declare corresponding permissions. Undeclared capabilities reduce transparency and informed consent, making it easier for a user or orchestrator to run a skill without understanding that local data may be accessed or artifacts may be created.

Tp4

High
Category
MCP Tool Poisoning
Confidence
96% confidence
Finding
The skill description says it crawls Wadiz/Tumblbug and generates DMs locally, but the analyzed behavior indicates Tumblbug is not actually supported and project information is sent to an external Anthropic API. This is dangerous because users may unknowingly disclose third-party scraped data or sensitive business prompts to an external service under a false assumption of local-only processing.

Missing User Warnings

Medium
Confidence
84% confidence
Finding
The skill omits a warning that it automates scraping of third-party platforms and generates outreach DMs to project makers. That omission increases legal, policy, and abuse risk because users are not prompted to consider platform terms, anti-spam rules, consent, or rate-limiting implications before using the tool.

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The code sends scraped project descriptions to Anthropic for DM generation whenever an API key is present, but it provides no explicit consent flow, disclosure, or opt-in about third-party data transfer. In this skill's context, the content comes from external project pages rather than the operator's own text, so sharing it with an LLM vendor can create privacy, compliance, or terms-of-service risk even if the data is publicly visible.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.