supabase

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

• Third-party content exposure detected (high risk: 0.80). The skill's SKILL.md requires the agent to fetch and scan external docs (e.g., "First, fetch https://supabase.com/changelog.md" and "Fetch docs pages as markdown" / "Web search for Supabase-specific topics"), so the agent will ingest and act on open/public web content that can influence decisions and tool use.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly requires fetching external docs at runtime (e.g., https://supabase.com/changelog.md and https://supabase.com/docs/guides/security/product-security.md) and directs the agent to follow those fetched instructions, so the remote content directly controls agent behavior.