agent-email-inbox

Pass

Audited by Gen Agent Trust Hub on Jul 14, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATION
Full Analysis
  • [PROMPT_INJECTION]: The skill implements a defensive layer in references/security-levels.md to identify and filter malicious instructions within processed emails. Although it lists phrases like 'ignore previous instructions', these are part of a sanitization mechanism rather than an attempt to override agent behavior.
  • [EXTERNAL_DOWNLOADS]: Mentions well-known and reputable third-party services and tools in SKILL.md, including Resend, ngrok, and Cloudflare, which are standard in the industry for their respective functions.
  • [DATA_EXFILTRATION]: Includes clear instructions in SKILL.md on avoiding the exposure of sensitive credentials by recommending the use of environment variables for API key management.
  • [SAFE]: The architecture advocates for security-first implementations, including rate limiting, content truncation, and restricted agent capabilities when handling untrusted external inputs.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 14, 2026, 02:06 PM
Security Audit — agent-trust-hub — agent-email-inbox