followup-sweep

Pass

Audited by Gen Agent Trust Hub on Aug 28, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources, creating a surface where malicious instructions embedded in emails could influence agent behavior.
  • Ingestion points: Reads received mail, sent mail, and calendar events via platform connectors (SKILL.md).
  • Boundary markers: Absent. The instructions do not specify any delimiters or safety prompts to ignore instructions that might be contained within the emails being analyzed.
  • Capability inventory: The agent can read sensitive user communications and create new Gmail drafts.
  • Sanitization: None identified. There is no mechanism described to sanitize or filter the content of processed emails before they are analyzed for "loops."
  • [NO_CODE]: The skill consists entirely of natural language instructions and configuration metadata without any accompanying executable scripts or binary files, reducing the risk of direct malicious code execution.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 28, 2026, 05:56 PM
Security Audit — agent-trust-hub — followup-sweep