The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes external data (user names, email addresses, and HR attributes) to construct and execute API requests for record creation in Salesforce, presenting a surface for indirect prompt injection.
Ingestion points: Personal and professional identity data is collected via user input as defined in the 'Required Inputs' section of SKILL.md.
Boundary markers: The skill implements a manual confirmation step in Phase 3, Step 5, where the agent must present the configuration plan and wait for user approval before proceeding with mutations.
Capability inventory: The skill leverages the mcp__headless-360__dispatch tool to perform POST operations creating User, Account, Employee2, and PermissionSetAssignment records, as outlined in references/mcp-invocation.md.
Sanitization: The instructions do not specify validation or sanitization requirements for user-provided strings before they are interpolated into the JSON bodies for the Salesforce REST API calls.

service-itsm-agentic-setup-uel-user-create