commerce-app-business-config
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONDYNAMIC_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests user-provided metadata—such as field names, labels, and descriptions—and interpolates them into the
app.commerce.config.tsconfiguration file. This creates an attack surface where maliciously crafted user input could potentially influence the resulting source code or build process. - Ingestion points: User-supplied 'Name', 'Type', 'Label', 'Description', and 'Default value' defined in Step 1 and written to
app.commerce.config.ts. - Boundary markers: The skill does not explicitly define boundary markers in the file generation step, relying on agent instructions.
- Capability inventory: The skill performs file writes and executes build commands (
aio app build). - Sanitization: Step 2 outlines validation constraints for various data types and enforces empty defaults for sensitive password fields to prevent credential exposure.
- [COMMAND_EXECUTION]: The skill executes
npx @adobe/aio-commerce-lib-app initto initialize project files andaio app buildto validate the generated configuration. These commands utilize official Adobe Commerce developer tools. - [DYNAMIC_EXECUTION]: The skill generates TypeScript code for
dynamicListfields, which includes factory functions that are executed at runtime to resolve configuration options from external sources. - [EXTERNAL_DOWNLOADS]: The skill uses
npxto download and execute the@adobe/aio-commerce-lib-appinitialization utility from the official npm registry.
Audit Metadata