etsy-tony-full-original-product

Pass

Audited by Gen Agent Trust Hub on Sep 17, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze data from public Etsy product pages provided by the user. This external content, which includes product descriptions, seller claims, and gallery information, is processed by the agent to generate new product concepts. Because this data originates from an untrusted third-party source (Etsy sellers), it could potentially contain malicious instructions intended to influence the agent's behavior.
  • Ingestion points: references/original-development-workflow.md (Step 1: Capture the Source) describes opening public Etsy URLs and saving visible product data.
  • Boundary markers: The skill attempts to mitigate risks by categorizing data into visible_fact, seller_claim, and inference, and by creating a source-audit.md to separate inspiration from expression.
  • Capability inventory: The skill utilizes the built-in image_gen tool and writes various audit and brief files to the local file system (e.g., ./etsy-output).
  • Sanitization: The skill lacks programmatic sanitization of the fetched HTML/text, relying instead on structural instructions for the LLM to ignore distinctive source expressions.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 17, 2026, 03:46 AM
Security Audit — agent-trust-hub — etsy-tony-full-original-product