data-capture-sdk

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

• Third-party content exposure detected (high risk: 0.80). The skill's SKILL.md Live Data Sources explicitly instruct the agent to fetch and read public files from raw.githubusercontent.com (e.g., https://raw.githubusercontent.com/Scandit/data-capture-documentation/...), meaning it ingests open third-party web content at runtime which can directly influence product-selection decisions.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly instructs runtime fetches from raw GitHub URLs (e.g. https://raw.githubusercontent.com/Scandit/data-capture-documentation/main/src/data/products.json and the related raw.githubusercontent.com feature/symbology/system-requirements MDX URLs), and uses that fetched content to determine product recommendations and platform/feature availability, so external content directly controls the agent's outputs.