service-itsm-agentic-setup-itsm-agentforce-permset-assign

Pass

Audited by Gen Agent Trust Hub on Aug 27, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses the Bash tool to execute Salesforce CLI (sf) commands. These commands are used for querying org metadata, checking user identity, and assigning permission sets or licenses. All write operations are preceded by a mandatory user confirmation step via AskUserQuestion.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external data returned from Salesforce API calls and CLI queries. To mitigate risks associated with unstructured data, the skill routes all tool outputs through local Node.js scripts (scripts/*.mjs). These scripts parse the JSON payloads and perform deterministic validation, such as verifying User ID formats via regex and matching permission set names against a hardcoded allowlist, before returning simplified results to the agent's context.
  • [PRIVILEGE_ESCALATION]: While the skill's primary function is the assignment of permission sets and licenses, this behavior is documented as the intended purpose. The instructions strictly enforce least-privilege principles by defaulting to the running user and requiring explicit selection of a single persona rather than auto-assigning multiple permissions.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 27, 2026, 04:35 PM
Security Audit — agent-trust-hub — service-itsm-agentic-setup-itsm-agentforce-permset-assign