system-bus

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

• Third-party content exposure detected (high risk: 0.80). The SKILL.md explicitly registers and ingests incoming webhooks from third-party services (see src/webhooks/server.ts and src/webhooks/providers/* for Front, GitHub, Vercel, Todoist, Mux) which are untrusted/user-generated payloads that are processed by Inngest functions (src/inngest/functions/...) and can be fed into the LLM inference utility (src/lib/inference.ts) or trigger gateway actions, so external content can materially influence decisions and tool use.