service-email-to-case-configure
Pass
Audited by Gen Agent Trust Hub on Aug 28, 2026
Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill invokes the Salesforce CLI (
sf) usingsubprocess.runto handle org identification, session token retrieval, and SOQL data queries. - Evidence: Observed in
scripts/apply-casesettings.py(e.g.,sf_jsonfunction) andscripts/check-agent-email-capability.sh. - [DYNAMIC_EXECUTION]: The test infrastructure uses
importlib.utilto load and execute the skill's Python scripts (which contain hyphens in their filenames) as modules for automated testing. - Evidence: Located in
scripts/tests/_bootstrap.py. - [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied data such as routing names, email addresses, and usernames, creating a surface for potential injection if inputs are not strictly validated.
- Ingestion points: Command-line arguments (
--owner-value,--routing-email, etc.) inscripts/apply-casesettings.py. - Boundary markers: Not present.
- Capability inventory: Shell command execution, network SOAP requests, and file system writes.
- Sanitization: The script implements basic character escaping (single quotes and backslashes) for values used in SOQL queries to mitigate injection risks.
Audit Metadata