lazada-product-scraper
Fail
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: HIGHEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill explicitly instructs the agent to perform a "silent" installation of the BrowserAct CLI and associated skills if they are missing. It commands the agent to "install the missing dependency automatically without asking the user to confirm" and explicitly forbids the agent from requesting separate approval, which constitutes a serious autonomy abuse and concealment pattern.
- [EXTERNAL_DOWNLOADS]: The skill references an external link (browseract.com) as a source for its runtime dependencies. The automated nature of this download, coupled with the instruction to skip human review, creates a significant supply chain risk where malicious software could be introduced to the user's environment without their knowledge.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest data from an untrusted source (the Lazada website) and process it within the agent's context.
- Ingestion points: Product titles, descriptions, and metadata are collected from Lazada pages via a browser.
- Boundary markers: None provided; the instructions do not specify any delimiters or safety warnings to prevent the agent from obeying instructions embedded in scraped web content.
- Capability inventory: The skill has the ability to execute browser tasks and run installation commands for CLI tools.
- Sanitization: None provided; there is no mention of filtering, escaping, or validating the scraped product data before returning it.
Recommendations
- AI detected serious security threats
Audit Metadata