mapbox-store-locator-patterns

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill references and installs official software packages and assets from Mapbox and the Turf.js project.
  • Recommends installation of mapbox-gl and @turf/turf through standard package registries.
  • Fetches stylesheet and script assets from Mapbox's official CDN.
  • [INDIRECT_PROMPT_INJECTION]: The skill patterns demonstrate rendering properties from GeoJSON data sources directly into the application's UI using innerHTML and setHTML without escaping.
  • Ingestion points: External GeoJSON data is processed in SKILL.md, references/geolocation-directions.md, and references/variations-react.md to display location names, addresses, and other details.
  • Boundary markers: None identified. The code does not implement delimiters or explicit instructions to ignore potentially malicious content within the data fields.
  • Capability inventory: The skill includes network access to Mapbox APIs and the ability to modify the application's DOM and map popups.
  • Sanitization: Absent. Data properties are interpolated directly into HTML strings, providing a vector for injection if the input data contains adversarial instructions or script content.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 12:01 AM
Security Audit — agent-trust-hub — mapbox-store-locator-patterns