skills/brucesongs/kali-claw/web-xss/Gen Agent Trust Hub

web-xss

Fail

Audited by Gen Agent Trust Hub on Sep 19, 2026

Risk Level: CRITICALREMOTE_CODE_EXECUTIONDATA_EXFILTRATIONCOMMAND_EXECUTIONOBFUSCATIONPRIVILEGE_ESCALATIONPERSISTENCEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill contains functional examples of Remote Code Execution (RCE) and reverse shell payloads. Specifically, guides/xss-to-rce-escalation-guide.md includes a Node.js reverse shell using child_process.spawn and net.Socket, while guides/ssti-attack-guide.md provides a Jinja2-based reverse shell using os.popen. These scripts are designed to establish persistent command-line access to a victim server or machine.
  • [DATA_EXFILTRATION]: Numerous files, including payloads.md and guides/xss-to-rce-escalation-guide.md, provide functional JavaScript payloads for stealing sensitive data. Examples include exfiltrating session cookies (document.cookie), local storage tokens, and full page content to attacker-controlled domains like attacker.com and evil.com using fetch, navigator.sendBeacon, and the image source technique.
  • [OBFUSCATION]: The skill documents extensive methods for bypassing security filters through obfuscation. guides/waf-bypass-xss-guide.md and payloads.md detail techniques such as Base64 encoding of executable strings, hex-escaped JavaScript identifiers, and multi-layer URL encoding designed to evade signature-based detection in Web Application Firewalls (WAFs).
  • [COMMAND_EXECUTION]: In the context of Server-Side Template Injection (SSTI), the skill provides instructions in guides/ssti-attack-guide.md for executing arbitrary shell commands through the template rendering engine, leveraging internal system calls like subprocess.Popen and os.popen to manipulate the host operating system.
  • [PRIVILEGE_ESCALATION]: The guide guides/xss-to-rce-escalation-guide.md outlines methods for privilege escalation within Electron-based desktop applications. It describes how to escape renderer sandboxes, abuse nodeIntegration, and exploit contextIsolation misconfigurations to gain full system access from an initial XSS vulnerability.
  • [PERSISTENCE]: The skill includes instructions for maintaining persistence after an initial attack. guides/xss-to-rce-escalation-guide.md demonstrates how to register malicious Service Workers that survive page reloads and cache clearing, allowing for continuous data exfiltration and request interception.
  • [INDIRECT_PROMPT_INJECTION]: The skill is configured to use the WebFetch tool to process arbitrary external web content entering the agent's context. This creates an attack surface for indirect prompt injection, as there are no explicit boundary markers or sanitization logic defined in the skill's instructions to prevent embedded malicious instructions in fetched data from influencing the agent's behavior.
  • [DYNAMIC_EXECUTION]: The provided payloads frequently utilize dynamic execution sinks such as eval(), Function(), and setTimeout() with string arguments to execute dynamically constructed or decoded code at runtime, often to bypass static security analysis.
Recommendations
  • CRITICAL: 1 infected file(s) detected - DO NOT USE
  • AI detected serious security threats
  • Contains 4 malicious URL(s) - DO NOT USE
Audit Metadata
Risk Level
CRITICAL
Analyzed
Sep 19, 2026, 04:36 PM