sf-einstein

Pass

Audited by Gen Agent Trust Hub on Jun 29, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides conceptual architecture, configuration guidance, and best practices for Salesforce Commerce Cloud Einstein AI features including Recommendations, Predictive Sort, and Search Ranking.
  • [SAFE]: No hardcoded credentials, sensitive file access, or suspicious command execution patterns were identified.
  • [SAFE]: Data ingestion is limited to retrieving documentation from official or search-based sources (Salesforce documentation), which is consistent with the skill's intended purpose as a developer aid.
  • [SAFE]: The skill includes a standard ingestion surface for documentation retrieval.
  • Ingestion points: External content is ingested via web searches and fetches for Salesforce API documentation and Data Cloud references as specified in SKILL.md.
  • Boundary markers: Explicit boundary markers are not present in the fetch instructions, but the usage context is limited to developer documentation.
  • Capability inventory: The skill allows use of tools like Write, Edit, and Bash for implementing the AI features.
  • Sanitization: Content is fetched for reference; no explicit sanitization logic is instructed as the source is expected to be technical documentation.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 29, 2026, 02:51 PM
Security Audit — agent-trust-hub — sf-einstein