brand-naming

Pass

Audited by Gen Agent Trust Hub on May 19, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is entirely instructional and provides a structured methodology for brand naming. It does not contain executable code, remote script downloads, or requests for sensitive system access.
  • [PROMPT_INJECTION]: An analysis for indirect prompt injection surfaces shows that the skill ingests user-provided 'Brief' data (Step 1) which is subsequently used to interact with external tools like search engines and translation services (Step 4). While this represents a standard functional surface, there are no instructions to bypass safety filters or override agent behavior.
  • Ingestion points: User-provided naming brief in SKILL.md (Step 1).
  • Boundary markers: Absent.
  • Capability inventory: Search engine interaction (SERP/Google) and translation tool usage (Google Translate).
  • Sanitization: Absent.
  • [DATA_EXFILTRATION]: The skill's mentions of external services (USPTO TESS, EUIPO, Google) are limited to manual or tool-assisted screening of brand names for trademark and domain availability, which is consistent with the primary purpose of the skill.
Audit Metadata
Risk Level
SAFE
Analyzed
May 19, 2026, 03:58 PM
Security Audit — agent-trust-hub — brand-naming