send-weekly-update
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection because it ingests untrusted external data and uses it to perform actions like drafting or sending emails.
- Ingestion points: The skill reads from 'source sheet, report, notes, metrics, or communications' in SKILL.md.
- Capability inventory: The skill is capable of drafting and sending emails, including those with HTML content.
- Boundary markers: The prompt does not specify the use of delimiters or 'ignore embedded instructions' warnings for the data it processes from external sheets or notes.
- Sanitization: The skill lacks explicit instructions to sanitize or escape the content from source materials before interpolating it into the email body or HTML table.
Audit Metadata