location-ar-experience
Pass
Audited by Gen Agent Trust Hub on Sep 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines a data model (ARContent) and a loader (SpatialContentLoader) that ingest content from external APIs. This content includes natural language fields like title and description, which could be exploited to deliver malicious instructions to the agent.
- Ingestion points: SpatialContentLoader.update_position in SKILL.md fetches tile-based content from a remote API client.
- Boundary markers: The code snippets do not include delimiters or instructions to treat the fetched content as untrusted data.
- Capability inventory: The skill performs network requests and interacts with device location services.
- Sanitization: There is no evidence of validation or sanitization for the metadata fields retrieved from the API.
- [EXTERNAL_DOWNLOADS]: The skill documentation in references/arcore-geospatial.md specifies a dependency on the com.google.ar:core library. This is a standard development package from a well-known and established service provider.
Audit Metadata