service-omni-skills-based-routing-configure
Pass
Audited by Gen Agent Trust Hub on Sep 6, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied agent lists and skill names which are interpolated into SOQL queries and Salesforce metadata. This ingestion surface is secured through strict regex validation of identifiers and XML entity escaping of labels before processing. Evidence: agent_members_csv and skill_developer_name inputs in scripts/configure-and-report.sh.\n- [COMMAND_EXECUTION]: The script performs Salesforce CLI (sf) operations to query org data, deploy metadata, and execute REST API calls. These commands are constructed using validated parameters and are restricted to the user-provided org alias. Evidence: sf commands in scripts/configure-and-report.sh.\n- [DYNAMIC_EXECUTION]: The skill includes a testing harness in scripts/tests/_bootstrap.py that uses subprocess.run to execute the main configuration script for automated validation. This is standard functionality for skill maintenance and contract testing.
Audit Metadata