Skills
skills/agentmail-to/agentmail-skills/agentmail-mcp/Gen Agent Trust Hub

agentmail-mcp

Pass

Audited by Gen Agent Trust Hub on Apr 17, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: Potential for indirect prompt injection via email content processing.
  • Ingestion points: Untrusted data enters the agent context through the list_threads and get_thread tools in SKILL.md.
  • Boundary markers: The instructions contain no delimiters or warnings to ignore instructions embedded in email bodies.
  • Capability inventory: The skill has access to impactful tools including send_message, reply_to_message, forward_message, and delete_inbox.
  • Sanitization: There is no description of content validation or escaping for processed emails.
  • [EXTERNAL_DOWNLOADS]: Installs the agentmail-mcp package from the npm and pip registries, which are standard for the vendor's tools.
  • [COMMAND_EXECUTION]: Provides instructions to run the MCP server locally via npx -y agentmail-mcp or by executing Python binaries.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 17, 2026, 05:18 PM
Security Audit — agent-trust-hub — agentmail-mcp