email-check

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingest untrusted data from external emails to perform automated triage and generate replies, which could allow an attacker to influence agent behavior through malicious email content.
  • Ingestion points: The workflow fetches unread messages via tools like search_gmail_messages and get-mail-message (referenced in SKILL.md sections 2 and 3).
  • Boundary markers: The instructions do not define clear delimiters or provide the agent with "ignore embedded instructions" warnings when processing the bodies of incoming emails.
  • Capability inventory: The skill uses filesystem tools (Read, Grep, Glob) and email MCP tools to analyze content and generate suggestions based on a local email-profile.md file.
  • Sanitization: No sanitization or validation logic is present to filter out prompt injection attempts hidden within email text before the agent processes them for scoring or drafting.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 10:38 AM
Security Audit — agent-trust-hub — email-check