creativeclaw-product-photoshoot

Pass

Audited by Gen Agent Trust Hub on Sep 6, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data such as product references and logos via the search_assets tool and manual uploads. This creates a surface where instructions hidden in assets could potentially influence the agent. This ingestion is necessary for the skill's primary task of image consistency.\n
  • Ingestion points: Asset search and manual upload workflows in SKILL.md.\n
  • Boundary markers: No specific delimiters are defined to separate asset data from system instructions.\n
  • Capability inventory: The skill integrates with tools for asset management, theme retrieval, and image generation using multiple models.\n
  • Sanitization: No explicit sanitization or validation of asset content is described.\n- [EXTERNAL_DOWNLOADS]: The skill connects to a vendor-owned MCP tool at https://app.creativeclaw.co/mcp/chatgpt to manage media and feedback. This is a standard functional integration for the developer's infrastructure.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 6, 2026, 12:38 PM
Security Audit — agent-trust-hub — creativeclaw-product-photoshoot