supabase

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

• Third-party content exposure detected (high risk: 0.85). The skill’s runtime workflow fetches outsider-authored public web content (e.g., https://supabase.com/changelog.md and other Supabase docs pages as markdown) and feeds the fetched text into the agent’s LLM context for summarization/decision-making.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill explicitly requires fetching external markdown at runtime (e.g., https://supabase.com/changelog.md and other docs fetched by appending .md such as https://supabase.com/docs/guides/security/product-security.md), and those fetched documents are used to drive the agent's instructions/behavior, so this is a runtime dependency that directly controls prompts.