competitor-profiling

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

• Third-party content exposure detected (high risk: 0.95). SKILL.md requires runtime scraping of outsider-authored competitor web pages (Firecrawl map/scrape/search) and review pages, whose extracted markdown/body text is then ingested into the agent’s LLM context for synthesis.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill explicitly scrapes user-supplied competitor sites at runtime (e.g., https://competitor1.com) via Firecrawl and pulls DataForSEO results, and that fetched content is injected into the profile synthesis (i.e., directly controls the model's output), so these runtime fetches are a dependency that influences prompts/behavior.