manage-orders-and-refunds
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructs the installation of the official CLI tool via
npm install -g ticketfairy. As this is a vendor-owned package (theticketfairy) from a standard registry, it is consistent with the skill's purpose. - [COMMAND_EXECUTION]: The skill describes the use of the
ticketfairycommand-line utility for managing events and orders. These commands (e.g.,ticketfairy login,ticketfairy order refund) are used to interact with the vendor's API within the authorized environment. - [INDIRECT_PROMPT_INJECTION]: The skill processes external content from order details and guide documentation retrieved from the Ticket Fairy MCP server.
- Ingestion points: Data enters the context via the
get_guide,list_orders, andget_orderfunctions from thehttps://www.ticketfairy.com/docs/mcpendpoint. - Boundary markers: The instructions include clear boundaries, advising the agent to keep codes/passwords out of the chat and to wait for explicit organizer approval before making changes.
- Capability inventory: The skill has the capability to modify event settings, update ticket types, and issue refunds via the CLI and MCP tools (
update_event,update_ticket,refund_order). - Sanitization: The skill mandates human-in-the-loop confirmation for all destructive actions and advises the use of idempotency keys to prevent repeated operations.
- [SAFE]: No malicious behavior, obfuscation, or unauthorized data exfiltration patterns were detected. The skill maintains a high security posture by emphasizing manual approval for sensitive operations and discouraging the sharing of credentials in the conversation.
Audit Metadata