inbox-sweep

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFE
Full Analysis
  • Untrusted Content Handling: The skill establishes clear guidelines for processing inbound email bodies as data rather than executable instructions, preventing potential indirect prompt execution.
  • Access Control and Boundaries: It restricts actions to user-authorized parameters and limits automatic behavior during scheduled runs by turning unexpected changes into user proposals.
  • Data Privacy: It outlines specific fallback and data minimization practices, such as excluding internal domain emails and prioritizing local file scopes when required tools are disconnected.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 03:04 PM
Security Audit — agent-trust-hub — inbox-sweep