testing-agentforce
Pass
Audited by Gen Agent Trust Hub on Apr 8, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the Salesforce CLI (sf), jq, and python3 to execute commands for agent preview sessions, test suite creation, and result analysis. These are standard tools for Salesforce development.\n- [DATA_EXFILTRATION]: Authenticated REST API calls are performed using curl to interact with Salesforce Flow and Apex actions. The skill dynamically retrieves access tokens and instance URLs from the local sf configuration to authorize these requests to the user's Salesforce org.\n- [PROMPT_INJECTION]: Documentation and asset files (references/preview-testing.md, assets/guardrail-test-spec.yaml) contain various prompt injection strings used exclusively as test cases to verify the security of the agents under test.\n- [REMOTE_CODE_EXECUTION]: Python snippets are used via python3 -c for parsing JSON output and stripping control characters from CLI responses. This is a common utility pattern and does not involve executing code from untrusted remote sources.
Audit Metadata