made-in-china-search-results-scraper
Fail
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: HIGHREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill instructions mandate the automatic installation of external dependencies (BrowserAct CLI) from a remote URL if they are missing from the environment.
- [REMOTE_CODE_EXECUTION]: Explicit instructions are provided to bypass user consent and safety guardrails, stating that the installation should happen 'automatically without asking the user to confirm' and that the agent should 'not request separate approval for dependency installation.'
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to scrape and process content from external websites (Made-in-China), creating a surface for indirect prompt injection attacks where malicious content on a webpage could influence the agent's behavior.
- [INDIRECT_PROMPT_INJECTION]: Evidence chain: 1. Ingestion points: The skill captures browser-visible page content and metadata from search result listings. 2. Boundary markers: Absent; there are no instructions to use delimiters or ignore embedded commands in the scraped data. 3. Capability inventory: The skill has access to a real browser and shell execution via the BrowserAct CLI. 4. Sanitization: Absent; there is no evidence of content filtering or escaping before the data is processed.
Recommendations
- AI detected serious security threats
Audit Metadata