data-capture-sdk

Pass

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill fetches live product availability, feature matrices, and system requirements from the author's official GitHub repository (Scandit/data-capture-documentation). These resources ensure the agent provides current information on SDK capabilities and platform support.
  • [COMMAND_EXECUTION]: The skill instructs the agent to check for specific framework markers like package.json, pubspec.yaml, and Podfile in the user's project directory. This read-only check is used to automatically identify the development environment for tailored advice.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes project configuration files and remote documentation to provide context-aware recommendations. While this involves external data ingestion, the sources are either standard project manifests or trusted resources controlled by the vendor.
  • Ingestion points: Project metadata files (e.g., package.json, build.gradle) and remote documentation files from the Scandit GitHub organization.
  • Boundary markers: None explicitly defined for remote documentation content.
  • Capability inventory: The skill is advisory and does not have write access to the filesystem or execution capabilities for remote code.
  • Sanitization: Data is retrieved from authoritative vendor sources to provide accurate product information.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 14, 2026, 04:34 PM
Security Audit — agent-trust-hub — data-capture-sdk