service-itsm-agentic-setup-cmdb-access-assign
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill's primary function is the administrative management of user permissions in a Salesforce org. It follows security best practices by implementing access checks, requiring user confirmation for write operations, and avoiding the handling of raw credentials.
- [COMMAND_EXECUTION]: The skill utilizes the
headless-360MCP server (a tool associated with the vendor 'forcedotcom') to execute REST API calls against Salesforce. These operations are scoped to permission management and are mediated by the MCP server's security model. - [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external sources, specifically Salesforce org records (User, PermissionSet). It mitigates risks by instructing the agent to resolve users uniquely, verify existing assignments, and escape single quotes in user-supplied values for SOQL queries. The risk associated with this ingestion surface is considered safe for the intended use case.
- [PRIVILEGE_ESCALATION]: While the skill assigns permissions, it is explicitly restricted to a predefined set of standard CMDB permission sets. It includes instructions to never create or edit permission sets, ensuring the skill cannot be used to grant arbitrary or unauthorized access beyond its documented scope.
Audit Metadata