matrixscan-count-android
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill integrates with Scandit's barcode recognition systems, which ingest untrusted external data (barcode payloads) into the application context.
- Ingestion points: Untrusted barcode data enters the agent context via
session.recognizedBarcodesandtrackedBarcode.barcode.datawithin listener callbacks defined inreferences/integration.mdandreferences/highlights.md. - Boundary markers: The instructions and provided code snippets do not include explicit boundary markers or warnings to the LLM to treat barcode content as untrusted input.
- Capability inventory: The skill primarily demonstrates UI-related capabilities (e.g.,
setContentView,addView,runOnUiThread) and basic data logging. It does not provide code for shell command execution, file system modification, or network transmission using the barcode data. - Sanitization: No sanitization, validation, or escaping logic is implemented for the barcode data before it is used or logged.
Audit Metadata