astro-forms

Pass

Audited by Gen Agent Trust Hub on Feb 20, 2026

Risk Level: SAFEDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [DATA_EXFILTRATION] (LOW): The skill performs network operations using fetch to several non-whitelisted domains including api.resend.com, api.brevo.com, script.google.com, and challenges.cloudflare.com. While these are legitimate service endpoints for the skill's stated purpose, they fall outside the predefined whitelist for automated network operations.\n- [PROMPT_INJECTION] (LOW): Detected an Indirect Prompt Injection surface where external data is processed and potentially influences downstream systems.\n
  • Ingestion points: Untrusted user input enters via form submissions as defined in references/schemas.md.\n
  • Boundary markers: No specific delimiters or boundary markers are used to wrap data interpolated into email templates.\n
  • Capability inventory: The skill possesses network capabilities to send data via assets/boilerplate/lib/email/resend.ts and assets/boilerplate/lib/sheets.ts.\n
  • Sanitization: Although structural validation is performed via Zod, the skill does not explicitly sanitize the content (e.g., HTML escaping) before it is embedded in HTML email templates in references/resend-setup.md, which could lead to injection issues if the recipient or an automated process interprets the payload.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 20, 2026, 05:50 AM