matrixscan-count-kmp
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of external data through barcode scanning results, which is an inherent part of its functionality. \n
- Ingestion points: Barcode data is retrieved through the
BarcodeCountSession.recognizedBarcodesproperty as described inreferences/integration.md. \n - Boundary markers: No special delimiters are implemented to separate scanned barcode strings from the agent's logic, which is standard for scanning applications. \n
- Capability inventory: The documented capabilities are limited to console logging and UI display updates. No high-risk operations like unauthorized file system access or external data exfiltration are demonstrated or encouraged. \n
- Sanitization: The instructions do not specify sanitization for the scanned data, but given the limited capabilities, this remains a safe surface for the intended use case. \n- [SAFE]: The skill enforces security best practices for secret management by requiring license keys to be masked in the chat interface and prohibiting the agent from searching the local project environment for credentials. All external references and dependencies are directed to official, trusted vendor domains and package registries.
Audit Metadata