learnbase-api

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

• Third-party content exposure detected (high risk: 0.80). This skill explicitly fetches tenant-provided content via GET /v1/storefront/tenant (used throughout the frontend flow) — including aiAssistantSettings.customPrompt, tenant.customCss (in Layout.tsx it is injected via dangerouslySetInnerHTML), navLinks and module/document/embed URLs and discussions — which are untrusted/user-generated third‑party content that the agent is expected to read and that can materially alter assistant behavior or actions.

MEDIUM W009: Direct money access capability detected (payment gateways, crypto, banking).

• Direct money access detected (high risk: 1.00). The skill explicitly exposes checkout and payment collection functionality: it documents a POST /v1/storefront/checkout/session endpoint (with required idempotency), handles payment-related webhooks/events, and describes "payment collection" as a core feature. Those are specific API surfaces for initiating and managing monetary transactions (i.e., moving money), not merely generic tooling. Therefore it grants direct financial execution capability.