payload

Pass

Audited by Gen Agent Trust Hub on Aug 11, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill references standard initialization commands such as npx create-payload-app@latest and various official Payload CMS packages (e.g., @payloadcms/db-mongodb, @payloadcms/plugin-seo). These are from well-known and trusted sources.
  • [COMMAND_EXECUTION]: The skill provides standard development commands (npx, pnpm dev) and explains how to use the Payload Local API. It explicitly warns about the security implications of overrideAccess: true in the Local API, which is a significant security best practice.
  • [DATA_EXFILTRATION]: No unauthorized network activity or credential harvesting was found. The skill correctly instructs developers to use environment variables (process.env.PAYLOAD_SECRET, process.env.DATABASE_URL) for sensitive configuration, following industry standards.
  • [PROMPT_INJECTION]: The instructions focus purely on software development patterns and do not contain any language attempting to override agent safety protocols or bypass constraints.
  • [REMOTE_CODE_EXECUTION]: The skill does not contain any patterns for downloading and executing untrusted scripts. All code examples are standard TypeScript/JavaScript for use within a Next.js/Payload environment.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 11, 2026, 06:53 AM
Security Audit — agent-trust-hub — payload