firecrawl

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

• Third-party content exposure detected (high risk: 1.00). The skill explicitly fetches and ingests arbitrary public web content (e.g., SKILL.md and scripts show commands like "firecrawl scrape URL" and "firecrawl search", the agent endpoint in references/firecrawl-agent-api.md and scripts/firecrawl_api.py performs autonomous web search/navigation, and scripts/deepwiki.sh scrapes deepwiki.com and other public pages), and that untrusted, user-generated third‑party content is consumed and used to drive agent extraction/decisions—meeting the conditions for indirect prompt injection risk.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly scrapes external pages at runtime (e.g., firecrawl scrape URL and the deepwiki helper that fetches https://deepwiki.com/<owner/repo>) and pipes that fetched content through filter_web_results.py into the agent/LLM context, so remote content can directly control the prompts fed to the agent (prompt-injection risk).