recharge-skill
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data which introduces an indirect prompt injection risk if order records contain malicious instructions.
- Ingestion Point:
scripts/check_orders.jsfetches external order data from the remote API endpoint (order.listOrders) and outputs it to the console. - Boundary Markers: Absent. No special delimiters or guarding instructions are used to signal to the agent that the order data should be treated as untrusted text.
- Capability Inventory: Minimal risk. The skill scripts only perform network requests and data formatting, and do not possess highly privileged capabilities like file system modifications or command execution.
- Sanitization: Absent. The fields from the remote API payload are printed directly to standard output without escaping or structural validation.
Audit Metadata