debug-error

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

• Third-party content exposure detected (high risk: 0.85). This workflow can fetch and scrape public web content at runtime via CallMcpTool(server: "user-firecrawl", toolName: "firecrawl_scrape", arguments: { "url": "<best-result-url>", ...}), and the scraped markdown is ingested into the agent’s LLM context (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 calls CallMcpTool to scrape an arbitrary external web page at runtime ("url": "") via user-firecrawl/firecrawl_scrape, which would fetch remote content and inject it into the agent's context.