spacetimedb-procedures

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

• Third-party content exposure detected (high risk: 0.80). Yes — the skill's documentation and examples (notably the "Calling Reducers from Procedures" example in SKILL.md that calls ctx.http.fetch(url), the "HTTP Requests" section, and the "Example: Calling an External AI API") show procedures fetching and parsing arbitrary external URLs/OpenAI responses and then using those responses to drive reducer calls, database writes, and returned values, so untrusted third‑party content can directly influence agent actions.