newebpay-refund
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- [Data Exposure & Exfiltration] (SAFE): The skill handles sensitive financial credentials (Merchant ID, Hash Key, Hash IV). It follows industry best practices by using environment variables (e.g., NEWEBPAY_HASH_KEY) and performing encryption locally before transmission.
- [External Downloads] (SAFE): Network activity is directed to official NewebPay payment gateway domains (ccore.newebpay.com and core.newebpay.com). No unauthorized data exfiltration was detected.
- [Command Execution] (SAFE): While the Bash tool is enabled in the skill frontmatter, the instructions and examples focus on standard source code creation for refund processing rather than executing dangerous system commands.
- [Indirect Prompt Injection] (LOW): The skill has an ingestion surface through user arguments, but the risk is mitigated by its purpose as a development template.
- Ingestion points: $ARGUMENTS variable in SKILL.md used for refund type selection.
- Boundary markers: None present.
- Capability inventory: Write, Edit, and Bash tools are available to the agent.
- Sanitization: None provided in the code templates, but the usage context is restricted to logical branching.
Audit Metadata