graphql-architect

Pass

Audited by Gen Agent Trust Hub on Jun 14, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: No prompt injection patterns were detected. The instructions are professional and strictly focused on the domain of GraphQL architecture and schema design.
  • [DATA_EXFILTRATION]: No evidence of data exfiltration was found. Network references are limited to documentation (GitHub Pages) and official industry specifications (Apollo Dev specs).
  • [CREDENTIALS_UNSAFE]: No hardcoded credentials or secrets were found. Code examples use environment variable placeholders (e.g., process.env.JWT_SECRET) following security best practices.
  • [REMOTE_CODE_EXECUTION]: No remote code execution patterns or suspicious external script downloads were detected.
  • [EXTERNAL_DOWNLOADS]: The skill mentions common and well-known industry packages (e.g., apollo-server, dataloader, graphql-ws) but does not execute any untrusted external code.
  • [COMMAND_EXECUTION]: No suspicious shell command execution or dynamic context injection was found.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 14, 2026, 10:31 AM
Security Audit — agent-trust-hub — graphql-architect