form-attribution

Warn

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • EXTERNAL_DOWNLOADS (MEDIUM): The skill promotes the use of a remote script from an untrusted source (https://cdn.jsdelivr.net/npm/form-attribution@latest/dist/script.min.js). The use of the '@latest' tag is a security risk as it allows for automatic updates that could introduce malicious code into a website via a supply chain attack without any manual review.
  • PROMPT_INJECTION (LOW): The skill creates a surface for indirect prompt injection by processing untrusted external data and injecting it into the DOM.
  • Ingestion points: The library reads from URL search parameters (e.g., UTMs, click IDs) and document referrer strings.
  • Boundary markers: No delimiters or safety instructions are mentioned to prevent the agent or website from treating injected parameters as trusted commands.
  • Capability inventory: The script utilizes a MutationObserver to automatically detect and inject hidden input fields into all forms found on the page.
  • Sanitization: The documentation does not provide evidence of sanitization or validation of the URL parameters before they are injected into the HTML forms.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 17, 2026, 06:33 PM