agentforce-generate
Pass
Audited by Gen Agent Trust Hub on Sep 10, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill includes Python scripts (
discover.py,scaffold.py,org_describe.py) that use thesubprocessmodule to execute Salesforce CLI (sf) commands. These scripts are used to query Salesforce org metadata, validate action targets, and generate boilerplate code. The command execution is scoped to legitimate development workflows. - [SAFE]: The skill incorporates multiple safety-first design patterns. It enforces a 'hard gate' for user approval of Agent Specs before proceeding with implementation. It also includes a detailed
safety-review-reference.mdthat instructs the agent on how to audit generated content for responsible AI practices, such as disclosure of AI identity and prevention of PII collection. - [SAFE]: The 'concealment' pattern flagged by static analysis in
references/agent-design-and-spec-creation.mdrefers to instructional guidance for the AI to resolve design questions from available context to improve user experience. It does not attempt to hide malicious execution or bypass security controls.
Audit Metadata