service-omni-command-center-analyze
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The script
scripts/analyze.shexecutessfCLI commands to perform SOQL and Tooling API queries against the target Salesforce organization. This is the core functionality required for org analysis. - [COMMAND_EXECUTION]: The skill's test harness in
scripts/tests/_bootstrap.pyusessubprocess.run()to execute the main analysis script within a mocked environment for contract verification. - [INDIRECT_PROMPT_INJECTION]: The skill ingests data from a Salesforce org through tool outputs, creating a potential surface for indirect injection if the org data is malicious.
- Ingestion points: Data is ingested via
sf data querycalls inscripts/analyze.shand stored in JSON variables for analysis. - Boundary markers: The skill follows a strict output contract, emitting a final structured JSON object with specific fields.
- Capability inventory: The skill is limited to read operations (SOQL/Tooling) and data processing with
jq; it lacks write capabilities or external network operations beyond the Salesforce CLI's standard functionality. - Sanitization: Supervisor usernames and IDs are validated using regular expressions to prevent injection before being incorporated into query strings.
Audit Metadata