foreign-trade-research
Pass
Audited by Gen Agent Trust Hub on Aug 18, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process large volumes of untrusted data from external websites via the
web_search_advanced_exatool and Playwright browser. - Ingestion points: Competitive analysis, pricing data, and reviews are pulled from arbitrary external domains and platforms like Amazon, LinkedIn, and Trustpilot into the agent's context.
- Boundary markers: The instructions do not specify the use of delimiters or explicit boundary markers when processing search results or scraped content to prevent the model from following instructions found in external text.
- Capability inventory: The agent has the ability to perform network searches, automate browser sessions via Playwright, and write output to the local filesystem as a markdown report.
- Sanitization: There are no instructions for filtering, escaping, or sanitizing the content retrieved from the web before it is analyzed and summarized by the LLM.
- [EXTERNAL_DOWNLOADS]: The skill fetches data from multiple external sources and well-known services (e.g., Amazon, LinkedIn, Trustpilot) to perform its research functions. This behavior is consistent with the skill's primary stated purpose of market analysis.
- [COMMAND_EXECUTION]: The skill utilizes Playwright for "Browser Fallback" to access dynamic pages or sites requiring JavaScript rendering. This involves browser automation capabilities to navigate and interact with web content.
Audit Metadata