etsy-seller
Pass
Audited by Gen Agent Trust Hub on Nov 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill's primary function involves ingesting and processing untrusted data from multiple sources, including user-provided product details, pasted Search Query Report (SQR) tables, and competitor listing data scraped from Etsy search results during Phase 4B. This external content is interpolated into the agent's context to generate listing copy and update local database files. There are no explicit instruction boundaries or sanitization protocols mentioned to mitigate the risk of malicious instructions embedded within these inputs.
- Ingestion points: Phase 3 (User input, SQR data, CSV batch processing) and Phase 4B (Competitor SERP scrape results).
- Boundary markers: Absent. The instructions do not define delimiters or warnings for the agent to ignore embedded instructions in the processed data.
- Capability inventory: Local script execution via
python3, filesystem modifications viaEdit/Writetools, and network searches. - Sanitization: Absent. No filtering or escaping logic is described for external data before context interpolation.
- [COMMAND_EXECUTION]: The skill requires the agent to execute a local Python script (
scripts/bootstrap.py) upon every invocation (Phase 0) to ensure the local state directory (~/etsy-listings/) is initialized. It also provides a shell command (find) for the agent to locate the script. While the script performs benign operations such as directory creation and template copying from the skill's own package, the routine execution of local code is a notable capability.
Audit Metadata