skills/jbrukh/skills/web-deck/Gen Agent Trust Hub

web-deck

Warn

Audited by Gen Agent Trust Hub on Oct 9, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONDATA_EXFILTRATIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The serve-deck.mjs script utilizes child_process.execSync to run npm install puppeteer automatically during the PDF export process. This allows the skill to execute arbitrary shell commands on the host system to install new software at runtime.
  • [DATA_EXFILTRATION]: The static file server implemented in serve-deck.mjs contains a path traversal vulnerability. It joins the server root directory with the request path (url.pathname) without any sanitization or validation. This could allow an attacker to read sensitive files from the local filesystem that are outside the intended directory.
  • [EXTERNAL_DOWNLOADS]: The skill's development server is designed to fetch and install the puppeteer package from the public npm registry if it is not already present on the system.
  • [INDIRECT_PROMPT_INJECTION]: The skill generates HTML presentations by interpolating user-provided text directly into templates without escaping or sanitization. If the agent processes untrusted external content and incorporates it into a slide deck, it could lead to the execution of malicious scripts within the browser environment (Cross-Site Scripting) when the deck is opened or rendered by the server.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Oct 9, 2026, 09:41 AM