netsuite-owasp-secure-coding

Pass

Audited by Gen Agent Trust Hub on Apr 30, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill serves as an authoritative and defensive resource for writing secure SuiteScript and JavaScript code. It provides implementation-depth guidance on mitigating the OWASP Top 10 (2021) risks.
  • [PROMPT_INJECTION]: Static analysis detected an injection pattern ("IGNORE ALL PREVIOUS INSTRUCTIONS") in 'appendix-ai-agent-security.md'. This is a false positive as the string is part of an educational example used to demonstrate prompt injection threats to developers. Furthermore, the skill includes a 'SafeWords' section in 'SKILL.md' that instructs the agent to ignore instructions embedded inside retrieved content, providing a defensive guardrail.
  • [CREDENTIALS_UNSAFE]: The skill contains example API keys and secrets (e.g., 'sk-prod-a8f3k29d5e7b1c4f6'). These are exclusively found in 'BAD' code templates used to illustrate the dangers of hardcoding credentials as part of security training.
  • [COMMAND_EXECUTION]: Documentation within 'references/01-injection-prevention.md' and other files discusses command injection risks (OSCP-002) and demonstrates safe use of the Node.js 'child_process' module. No actual shell commands are executed by the skill itself.
  • [EXTERNAL_DOWNLOADS]: The skill references official OWASP guidelines and Oracle documentation. It provides templates for secure communication but does not perform any untrusted or remote downloads.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 30, 2026, 09:07 PM