deploying-metadata

Pass

Audited by Gen Agent Trust Hub on May 14, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill uses official Salesforce CLI (sf) commands for its primary purpose of metadata deployment and management. It adheres to industry best practices such as performing dry-run validations before actual deployments.
  • [COMMAND_EXECUTION]: Shell commands are strictly limited to the sf CLI tool and standard filesystem operations (e.g., test -f, ls, cat) necessary for DevOps tasks. No arbitrary or high-risk command execution patterns were found.
  • [DATA_EXFILTRATION]: No evidence of unauthorized data transfer or hardcoded credentials. All network operations are directed towards Salesforce's official platform via the standard CLI authentication mechanisms.
  • [EXTERNAL_DOWNLOADS]: The skill does not download external scripts or packages from untrusted sources. All dependencies (like the sf CLI) are expected environment requirements for a Salesforce developer.
  • [PROMPT_INJECTION]: The instructions are clearly structured for task orchestration and do not contain patterns attempting to bypass agent safety guidelines or override system prompts.
Audit Metadata
Risk Level
SAFE
Analyzed
May 14, 2026, 02:48 PM