Is Codex Agent Safe?

Codex Agent — Nerq Trust Score 48.2/100 (D grade). Score based on 2 independent trust signals. Last analyzed: 2026-03-25

Codex Agent is a npm package with a Nerq Trust Score of 48.2/100 (D), based on 3 independent data dimensions. Last analyzed: 2026-03-25 Security: 90/100. Popularity: 0/100. Data sourced from npm registry, GitHub repository, NVD, OSV.dev, and OpenSSF Scorecard. Last updated: 2026-03-25. Machine-readable data (JSON).

Safety Guide: Codex Agent

What is Codex Agent?

Codex Agent is a Node.js package — > Codex Agent 是对 OpenAI Codex CLI 的“守护与编排层”，补齐持续任务调度、异常自愈与多渠道通知能力，让 `codex exec --json` 可以安全地长时间运行并在断连后自动恢复。.

How to Verify Safety

Run npm audit to check for vulnerabilities. Review the package's GitHub repository for recent commits.

You can also check the trust score via API: GET /v1/preflight?target=codex-agent

Key Safety Concerns for Node.js package

When evaluating any Node.js package, watch for: dependency vulnerabilities, malicious packages, typosquatting.

Measured Signals

Codex Agent has a Nerq Trust Score of 48/100 (D). This score is a composite of automated measurements of security, maintenance, community, and quality signals.

Key Takeaways

• Codex Agent has a Trust Score of 48/100 (D).
• The score is a measured composite — it is not a suitability judgment. Evaluate the individual signals against your own requirements.
• Query the current measured values via the Nerq API.

Disclaimer: Nerq trust scores are automated measurements based on publicly available signals. They are not endorsements, verdicts, or guarantees of suitability. Always evaluate the signals against your own requirements.