Codex safeguards against accidental destructive actions
Melvin Vivas · X post · 2026-08-19 · Open on X
Topics: AI Safety, Security & Guardrails, AI Dev Tools & Productivity, AI Agents, Tool Use & MCP · Level: intermediate
Summary
The creator shares an update from Tibo at OpenAI about changes that lower the risk of Codex doing destructive things, like accidental deletions, while it works. He calls it a must-read for Codex users.
Key points
- OpenAI looked into reports of Codex performing potentially destructive actions.
- Over a few weeks they rolled out preventive fixes that reduce accidental deletions.
- Agent safety in coding tools is an active area to keep up with if you rely on autonomous agents.
Resources mentioned
- Tibo (@thsottiaux) · person · x.com · free
X account of the OpenAI Codex team member who posts Codex updates and usage-limit announcements.
Also in: Codex Tip: Ask Codex to Organize Your Recent Chats (Melvin Vivas on X · notes), An X research bot built with Hermes (Melvin Vivas on X · notes), Why Codex Usage Drains Faster: Prompt Cache Hit Rate (Melvin Vivas on X · notes), Run OpenAI Codex with Open-Source and Local Models (OSS Mode) (Melvin Vivas on X · notes) and 1 more - OpenAI Codex · tool · openai.com · paid · recommended by both Bashiri Smith & Melvin Vivas
OpenAI's coding agent. In the diagram it writes code, fixes review findings and drives the build loop. The creator also used it to make this video.
Also in: Developers moving from ChatGPT/Codex to Claude (Melvin Vivas on X · notes), Creator Opinion: Codex Is Still the Best at Computer Use (Melvin Vivas on X · notes), OpenAI Codex Ambassador applications are open (Melvin Vivas on X · notes), GPT-Live-1.1: Multilingual Voice Update for ChatGPT and Codex (Melvin Vivas on X · notes) and 264 more
Try this
- Read Tibo's recap on X of the Codex destructive-action safeguards.
More in AI Safety, Security & Guardrails
- OpenAI Astra: More Capable Cyber Model Than Sol
- Teaching Teens AI Safety: ChatGPT for Teens in Singapore
- Screening LLM Prompts and Responses on CPU with AIBackends + GliGuard
- Liquid AI's Multilingual PII Detection Works on Japanese
- Use a Local Model for Confidential Data with Your Agent
- Don't Let Coding Agents Run Database Migrations Unsupervised