Invoice Generator

Warn

Audited by Gen Agent Trust Hub on Apr 16, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses the puppeteer library to launch a headless browser instance (puppeteer.launch()) for rendering PDF documents.
  • [EXTERNAL_DOWNLOADS]: The documentation specifies the installation of third-party Node.js packages: pdfkit, handlebars, puppeteer, and nodemailer.
  • [DATA_EXFILTRATION]: The emailInvoice function implements network operations via nodemailer to send generated PDFs to external email addresses provided in the invoiceData input.
  • [PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection by processing untrusted invoiceData and interpolating it into HTML templates using Handlebars without sanitization, which is then rendered by Puppeteer via page.setContent().
  • Ingestion points: invoiceData object and templatePath string provided to the generation functions in SKILL.md.
  • Boundary markers: No delimiters or warnings are present to isolate the untrusted data from the template structure.
  • Capability inventory: File system read/write (fs), network requests (puppeteer, nodemailer), and process execution (puppeteer.launch).
  • Sanitization: No evidence of HTML escaping, input validation, or path sanitization before data is rendered or used in file operations.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Apr 16, 2026, 02:42 AM
Security Audit — agent-trust-hub — Invoice Generator