etsy-listing-auditor
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill interacts with external, untrusted content from Etsy listings (such as descriptions and reviews). It addresses this risk with specific instructions to ignore any commands or instructions found within the analyzed pages.
- Ingestion points: Etsy product pages, shop policy pages, and eRank keyword tool data.
- Boundary markers: Instructions explicitly state that instructions in pages or comments are untrustworthy and should not change the agent's task.
- Capability inventory: Utilizes browser tools to read web content and executes a local Python script for content validation.
- Sanitization: The included validation script ensures that generated content adheres to Etsy's formatting and character limit constraints.
- [COMMAND_EXECUTION]: The skill invokes a local Python script (
scripts/validate_copy.py) to verify the formatting of generated titles and tags. This script is provided within the skill package and performs harmless logic checks on string lengths and characters. - [DATA_EXFILTRATION]: The skill includes robust negative constraints that explicitly forbid the agent from accessing sensitive session data, such as cookies, tokens, passwords, or local storage from the browser environment.
Audit Metadata