platform-trial-org-create
Pass
Audited by Gen Agent Trust Hub on Aug 27, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill is authored by the platform vendor and utilizes standard tools (sf CLI, jq) to interact with Salesforce APIs. Analysis found no malicious patterns such as credential exfiltration, persistence mechanisms, or unauthorized remote code execution.
- [COMMAND_EXECUTION]: The bash scripts
create_signup_request.shandget_signup_request.shwrap CLI commands securely. The creation script includes a specificsf_quotefunction and validation logic that prevents parameter injection by blocking values containing conflicting quote types and ensuring all inputs are properly escaped for the CLI parser. - [INDIRECT_PROMPT_INJECTION]: The skill presents a surface for indirect prompt injection by reading data from an external API. 1. Ingestion points: The
scripts/get_signup_request.shscript retrievesSignupRequestrecord details from the Salesforce API. 2. Boundary markers: Data returned from the API is presented to the agent without specific delimiters or isolation instructions. 3. Capability inventory: The skill can execute bash scripts, interact with the Salesforce CLI, and write JSON artifacts to the workspace. 4. Sanitization: While input parameters are strictly sanitized for command execution, the data read back from the API is not filtered before display. This is a standard risk for management tools and is considered safe given the authenticated host org context.
Audit Metadata