studio-define-offer-review

Pass

Audited by Gen Agent Trust Hub on Aug 28, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests user-controlled content from local files (1-Product-Offer.md, 2-Customer-Persona.md) to perform its analysis, which represents a potential surface for indirect prompt injection if those files contain malicious instructions.
  • Ingestion points: SKILL.md (Workflow Step 1) instructions describe reading the Product Offer and Customer Persona files.
  • Boundary markers: The skill uses verbatim quoting for analysis but lacks explicit delimiters to ignore embedded instructions.
  • Capability inventory: The skill is limited to performing surgical file modifications via an Edit operation on specific local markdown files.
  • Sanitization: No explicit sanitization or filtering of external content is described.
  • [SAFE]: The skill implements secure directory traversal practices by explicitly forbidding searches in node_modules, build output, or vendored code directories, ensuring it operates only on intended project assets.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 28, 2026, 10:33 PM
Security Audit — agent-trust-hub — studio-define-offer-review