ClawHub
Back to skill

Security audit

talaria

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed browser stealth tool, but its broad WAF-evasion and scraping guidance plus automatic browser tooling downloads warrant user review before installation.

Install only if you need stealth browser automation for websites you are authorized to test or automate. Review the npm install/postinstall behavior, pin or audit dependencies in sensitive environments, and avoid putting proxy credentials in shared shells or logs.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Findings (7)

Lp3

Medium
Category
MCP Least Privilege
Confidence
84% confidence
Finding
The skill declares and relies on environment variables such as PROXY_SERVER, PROXY_USERNAME, PROXY_PASSWORD, and HEADLESS, but the finding indicates permissions are not explicitly declared in the skill's security model. That creates an authorization gap: the skill can influence network routing and consume secrets from the runtime environment without a clear permission boundary or user-facing approval model.

Vague Triggers

Medium
Confidence
79% confidence
Finding
The trigger and description are broad enough to activate on generic browser automation, scraping, bot detection, or Playwright issues, which can cause this stealth-evasion skill to be selected in situations where anti-detection behavior was neither necessary nor intended by the user. Because the skill is specifically designed to evade WAF and bot-detection signals, unintended invocation increases the chance of unauthorized or policy-violating automation.

Missing User Warnings

Medium
Confidence
90% confidence
Finding
Although the document says 'Only automate websites the user authorized access to,' the skill prominently markets stealth, anti-bot leak hiding, and WAF-trigger reduction without a strong, front-loaded warning about legal, policy, and abuse risks. In context, this is more dangerous than a normal browser skill because its core purpose is to reduce detection while interacting with protected sites, which can facilitate unauthorized scraping or access attempts.

Unpinned Dependencies

Low
Category
Supply Chain
Content
},
  "license": "MIT",
  "dependencies": {
    "playwright": "^1.49.0",
    "playwright-extra": "^4.3.6",
    "puppeteer-extra-plugin-stealth": "^2.11.2"
  }
Confidence
90% confidence
Finding
"playwright": "^1.49.0"

Unpinned Dependencies

Low
Category
Supply Chain
Content
"license": "MIT",
  "dependencies": {
    "playwright": "^1.49.0",
    "playwright-extra": "^4.3.6",
    "puppeteer-extra-plugin-stealth": "^2.11.2"
  }
}
Confidence
90% confidence
Finding
"playwright-extra": "^4.3.6"

Unpinned Dependencies

Low
Category
Supply Chain
Content
"dependencies": {
    "playwright": "^1.49.0",
    "playwright-extra": "^4.3.6",
    "puppeteer-extra-plugin-stealth": "^2.11.2"
  }
}
Confidence
91% confidence
Finding
"puppeteer-extra-plugin-stealth": "^2.11.2"

Known Vulnerable Dependency: playwright==1.49.0 — 1 advisory(ies): CVE-2025-59288 (Playwright downloads and installs browsers without verifying the authenticity of)

High
Category
Supply Chain
Confidence
97% confidence
Finding
playwright==1.49.0

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.