landing-page-intel
Pass
Audited by Gen Agent Trust Hub on Mar 24, 2026
Risk Level: SAFEDATA_EXFILTRATIONPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [DATA_EXFILTRATION]: The
scripts/scrape_landing_page.pyscript performs outbound network GET requests to fetch HTML content from arbitrary URLs provided as input. This behavior is the primary intended function of the scraping tool. - [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it retrieves and processes untrusted data from the internet (including HTML comments, meta descriptions, and CTA text) and includes it in the AI agent's context.
- Ingestion points: The
scrape_pagefunction inscripts/scrape_landing_page.pyingests untrusted content via therequestslibrary. - Boundary markers: Results are returned in JSON or a summary table format, but the skill does not provide explicit boundary markers or instructions to the agent to ignore commands within the scraped content.
- Capability inventory: The skill's scripts are limited to network fetching and regex-based parsing; they do not have permissions for file system writes, subprocess execution, or other high-privilege operations.
- Sanitization: The script uses regular expressions to isolate specific data points and employs standard JSON serialization for output, which provides structural isolation but does not sanitize the semantic content for natural language instructions.
- [EXTERNAL_DOWNLOADS]: The skill requires the
requestsPython package to be installed for handling HTTP communication.
Audit Metadata