codex-huge-context

Pass

Audited by Gen Agent Trust Hub on Aug 3, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill implements best practices for secret management by explicitly forbidding plaintext keys and environment variables, instead relying on the macOS Keychain via the /usr/bin/security tool.
  • [SAFE]: The preflight.rb script performs validation of local configuration files and ensures the authentication helper is present and functional. It includes explicit memory clearing of credentials to prevent leakage.
  • [SAFE]: The skill uses well-known and authoritative service endpoints (api.openai.com) and adheres to a well-defined deployment model for internal fleet maintenance.
  • [COMMAND_EXECUTION]: The skill executes shell commands for Keychain access (/usr/bin/security) and verification (codex login status, jq). These are standard administrative actions for the stated purpose of model configuration and do not pose a risk in this context.
  • [EXTERNAL_DOWNLOADS]: The skill references standard OpenAI API endpoints and official project repositories. No untrusted or suspicious external downloads were detected.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 3, 2026, 01:27 AM
Security Audit — agent-trust-hub — codex-huge-context