b2c-cap

Pass

Audited by Gen Agent Trust Hub on Jun 30, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSCREDENTIALS_UNSAFE
Full Analysis
  • [COMMAND_EXECUTION]: Uses the b2c CLI tool (or @salesforce/b2c-cli via npx) to perform commerce operations, including listing, pulling, validating, and installing packages on B2C Commerce instances.
  • [EXTERNAL_DOWNLOADS]: References the official @salesforce/b2c-cli Node.js package, which is a trusted vendor resource provided for Salesforce development.
  • [CREDENTIALS_UNSAFE]: The skill documents the use of the b2c setup inspect --unmask command, which reveals sensitive configuration details and secrets retrieved from environment variables or local files like dw.json and ~/.mobify. This is an intended feature of the vendor's CLI for configuration debugging and auditing.
  • [PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection by processing Commerce App Packages (CAPs) originating from external B2C instances or local file uploads.
  • Ingestion points: Remote B2C instances (via list, pull, and tasks commands) and local file system (via validate, package, and install commands).
  • Capability inventory: Shell command execution via the b2c CLI tool, which is capable of network requests and file system modifications.
  • Boundary markers: None described in the skill instructions; external content is processed directly by the CLI.
  • Sanitization: Relies on the CLI tool's internal validation logic for package content and structure.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 30, 2026, 02:57 PM
Security Audit — agent-trust-hub — b2c-cap