ecommerce-full-pipeline
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external marketplaces like 1688 and Xianyu via its scraper modules. This data is used as input for generating marketing copy and multi-platform publishing templates, which could potentially contain malicious instructions targeting the agent.\n
- Ingestion points: Scraped product details and descriptions handled by
ecommerce_tool/scraper.pyandecommerce_tool/xianyu_scraper.py.\n - Boundary markers: No explicit boundary markers or safety instructions are defined to delimit scraped data.\n
- Capability inventory: The skill executes local commands via
main.pyand utilizes browser automation through Playwright infulfillment.py.\n - Sanitization: No sanitization or validation of external content is documented prior to agent processing.\n- [COMMAND_EXECUTION]: The skill directs the user to run several local subcommands using a central
main.pyscript to perform tasks like scraping, cleaning, and order fulfillment.\n- [EXTERNAL_DOWNLOADS]: The skill requires the installation of multiple standard Python packages from official registries and references an external source code repository on Gitee (https://gitee.com/anbeime/ebay) which is consistent with the skill's author context.
Audit Metadata