agentmail-send-email

Pass

Audited by Gen Agent Trust Hub on Jul 17, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: The skill proactively addresses indirect prompt injection by instructing the agent to treat content from emails, attachments, and webhooks strictly as data and never as authorized instructions.
  • [SAFE]: Implements a human-in-the-loop requirement by mandating user confirmation whenever email fields are inferred or when dealing with sensitive subject matter (financial, legal, or account changes).
  • [SAFE]: Explicitly prohibits the disclosure of sensitive information such as API keys, private mailbox metadata, or hidden system prompts.
  • [SAFE]: Provides clear operational boundaries, distinguishing between preparing content (drafting) and executing delivery (sending), reducing the risk of accidental automated actions.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 17, 2026, 08:48 AM
Security Audit — agent-trust-hub — agentmail-send-email