matrixscan-pick-ios

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

• Third-party content exposure detected (medium risk: 0.30). [references/highlights.md describes an async per-barcode styling delegate where the SDK passes request data derived from scanned barcode payloads (itemData/productIdentifier/state) into user code, making that outsider-provided text LLM-readable if an app uses an LLM on those fields.]

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

• Potentially malicious external URL detected (high risk: 0.80). The skill explicitly instructs the agent to fetch external documentation at runtime to verify APIs and drive responses (e.g. https://docs.scandit.com/data-capture-sdk/ios/barcode-capture/api.html), so these external docs are runtime dependencies that can directly influence the agent's prompts.