web-crawler

Pass

Audited by Gen Agent Trust Hub on Oct 31, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill implements a secure authentication pattern by using a workspace proxy to inject credentials for ScrapeCreators, Firecrawl, and Apify. It explicitly instructs the agent not to search for environment variables or ask the user for API keys, which follows security best practices.
  • [SAFE]: All external network requests are routed to well-known technology services and official API endpoints (e.g., Google, Apple iTunes, Internet Archive, and Apify) as part of the tool's core scraping functionality.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted content from the web and social media, creating an attack surface for indirect prompt injection.
  • Ingestion points: Data is fetched from various third-party websites and social platforms via the exports.py script.
  • Boundary markers: No specific delimiters or "ignore previous instructions" warnings are applied to the scraped content before it is passed to the agent.
  • Capability inventory: The skill possesses network read capabilities and processes content into Markdown, but does not perform file-system writes or execute system commands based on the scraped input.
  • Sanitization: The tool performs basic HTML unescaping and regex-based whitespace normalization, but does not include specific logic to detect or neutralize embedded prompt injection instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 31, 2026, 04:04 PM