google-no-code
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process external, untrusted data from Google Form submissions, which could be used as a vector for indirect prompt injection.
- Ingestion points: The skill reads submission data using
e.response.getItemResponses()inSKILL.md. - Boundary markers: The code snippets do not use explicit delimiters or instructions to the LLM to ignore embedded commands within the form response data.
- Capability inventory: The provided script templates have the capability to send emails (
MailApp.sendEmail) and write data to spreadsheets (SpreadsheetApp.appendRow). - Sanitization: Example 2 in
SKILL.mdimplements basic sanitization for spreadsheet logging by prepending a single quote to prevent formula injection (e.g., values starting with=,+,@, or-).
Audit Metadata