temu-competitor-search
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted product information, such as titles and seller descriptions, from Temu search results. Maliciously crafted product data could potentially attempt to influence the agent's behavior during analysis.
- Ingestion points: The
scripts/scrape_temu.pyscript fetches and parses HTML content from the Temu website. - Boundary markers: The skill output is structured into a predefined JSON schema, which helps separate external data from agent instructions.
- Capability inventory: The skill utilizes network access via the
requestslibrary and executes local Python logic for data parsing. - Sanitization: Keywords are URL-encoded using
quote_plus, and the parsing logic is targeted toward specific HTML tags and classes. - [EXTERNAL_DOWNLOADS]: The skill fetches product data and search results from official Temu domains (
www.temu.com,www.temu.com/uk,www.temu.com/eu) to perform its primary function of price comparison. - [EXTERNAL_DOWNLOADS]: The
scripts/setup.shscript installs therequestsandbeautifulsoup4libraries from the official Python Package Index (PyPI) to support scraping capabilities.
Audit Metadata