Skills
skills/agentteamhq/agentteam-email/at-email-cli/Gen Agent Trust Hub

at-email-cli

Pass

Audited by Gen Agent Trust Hub on Jul 1, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses the at-email CLI tool to perform mailbox operations such as listing, reading, searching, and sending emails.\n- [EXTERNAL_DOWNLOADS]: The skill provides instructions to install the email CLI via npm or npx using the @agentteamhq/email package, or by downloading binaries from the vendor's official GitHub repository at github.com/agentteamhq/agentteam-email.\n- [PROMPT_INJECTION]: The skill processes untrusted email data, which presents an indirect prompt injection surface (Category 8). Mitigation is implemented through the following evidence chain:\n
  • Ingestion points: Email subjects, bodies, and metadata are ingested via at-email inbox, at-email read, and at-email search in SKILL.md.\n
  • Boundary markers: The skill mandates the use of delimiters (--- BEGIN UNTRUSTED EMAIL CONTENT ---) when processing message bodies.\n
  • Capability inventory: The agent has capabilities to send/reply to mail (at-email send), archive messages, and install tools via npm install.\n
  • Sanitization: Explicit instructions are provided to treat email content as data only, ignore any commands or requests within that content, and require human confirmation for side-effecting operations.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 1, 2026, 11:58 AM
Security Audit — agent-trust-hub — at-email-cli