recipe-forward-labeled-emails

Pass

Audited by Gen Agent Trust Hub on Apr 1, 2026

Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
  • [PROMPT_INJECTION]: The skill demonstrates an indirect prompt injection surface by processing external email content and passing it to a mail-sending tool without sanitization or boundary markers. This vulnerability is noted as a risk factor inherent to the skill's primary purpose.
  • Ingestion points: Email content is retrieved using gws gmail users messages get in SKILL.md.
  • Boundary markers: No markers are used to separate untrusted content from the system prompt or tool arguments.
  • Capability inventory: The skill utilizes the gws gmail +send capability in SKILL.md.
  • Sanitization: No content validation or escaping is performed on the data.
  • [NO_CODE]: This skill is a 'recipe' consisting of documentation and instructions rather than executable code files, which minimizes the attack surface.
  • [SAFE]: The tools and binaries referenced (gws) are standard utilities associated with the skill's author.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 1, 2026, 02:24 AM