recipe-batch-reply-to-emails

Pass

Audited by Gen Agent Trust Hub on Mar 4, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill uses the 'gws' (Google Workspace) command-line tool, which is consistent with the author 'googleworkspace'. All described operations—listing, reading, sending, and modifying emails—are standard and intended functions for this utility.
  • [PROMPT_INJECTION]: The skill processes untrusted external data (email content), creating a surface for indirect prompt injection.
  • Ingestion points: The skill retrieves email message content using the gws gmail users messages get command in SKILL.md.
  • Boundary markers: There are no specific delimiters or instructions provided to the agent to treat the ingested email content as data rather than instructions.
  • Capability inventory: The skill possesses the ability to send new emails (gws gmail +send) and modify message labels, which could be exploited if an ingested email contains malicious instructions that the agent follows.
  • Sanitization: The recipe does not specify any sanitization or validation of the email body before it is used by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 4, 2026, 10:50 PM