build-competitive-brief

Pass

Audited by Gen Agent Trust Hub on Jul 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • Dynamic HTML Verification: The workflow supports generating self-contained HTML files and optionally verifying their appearance using a browser-based renderer. This involves the dynamic execution of generated content as part of the report validation process.\n- Ingestion of External Data: The skill processes information from various sources including CRM systems, internal messaging, and public research. This activity creates a potential surface for indirect prompt injection, which is a common security consideration when agents aggregate and summarize content from external environments.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 15, 2026, 02:40 AM
Security Audit — agent-trust-hub — build-competitive-brief