ops-agent-skills

Pass

Audited by Gen Agent Trust Hub on Jun 9, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill defines legitimate operating procedures for business tasks. It emphasizes using existing systems and platform-managed integrations, reducing the risk of unauthorized tool setup.
  • [PROMPT_INJECTION]: The skill is designed to process data from external sources like email and support tools, creating a surface for indirect prompt injection.
  • Ingestion points: External data enters through Gmail/AgentMail (receipts, triage) and Intercom (customer support context).
  • Boundary markers: The instructions do not define explicit delimiters for ingested data.
  • Capability inventory: The agent can perform transactions via Stripe and update records in Google Sheets.
  • Sanitization: The skill includes natural language guardrails requiring a summary before irreversible or money-moving actions.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 9, 2026, 09:34 PM
Security Audit — agent-trust-hub — ops-agent-skills