firecrawl-scrape

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.95). The skill’s runtime path “scrape one or more URLs” fetches public web pages (outsider-authored content) and converts their body text into LLM-readable markdown/JSON context via Firecrawl’s scraping pipeline.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.80). The skill fetches arbitrary runtime URLs (e.g., the example https://example.com/pricing) and injects the scraped page content into an LLM query/response flow (the --query option and "LLM-optimized markdown"), so fetched remote content is used at runtime to control prompts/agent outputs.