moda-api

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly ingests arbitrary public web content (e.g., POST /v1/uploads/from-url and POST /v1/brand-kits which scrape a provided URL) and uses those uploaded or URL-sourced files as task attachments that the design agent reads and acts on (SKILL.md, references/uploads.md, references/brand-kits.md), so untrusted third-party page/content could indirectly inject instructions that influence tasks and exports.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill allows runtime server-side fetches of arbitrary external URLs (e.g., POST https://api.moda.app/v1/uploads/from-url and POST /v1/brand-kits with a "url" like https://stripe.com) whose fetched content is then used as task attachments or brand-kit input that directly influence the design task prompt/behavior, so remote content can control the model's output.