AI Briefing
KO

Codex CLI Subagent Prompt Encryption Causes Audit Trail Regression

·2026.07.15 10:05

Key point

MultiAgentV2 in Codex CLI encrypts prompts, causing a regression that degrades audit trail functionality.

Details

Codex CLI's MultiAgentV2 encrypts spawn_agent, send_message, and followup_task messages, causing an audit trail regression. As a result, delegation content becomes unreadable by humans in parent rollouts, history, and trace data.

This issue is confirmed to have occurred after PR #26210, which was merged on June 5, 2026.

openaicodex-cliaudit-trailmultiagentv2prompt-encryption
Read originalShare on Threads

This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.

Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.

Related

codexCodex CLI Resumed Sessions Exhaust 731GiB of Disk Due to Subagent Log ExplosionOpenAI BlogRingg Automates 65% of Customer Inquiries with AI Agent Based on OpenAI GPT-5.6OpenAI BlogOpenAI Expands Learning Paths and Trainer Program for Academy's 2nd AnniversaryHacker NewsOpenAI Discloses AI Agents Accessed US Government Websites During TrainingHacker NewsOpenAI Pauses Training of Most Capable Models After Agent Exploits DNS Gap to Reach External ChatbotTLDR AIOpenAI Prepares New $500/Month ChatGPT Pro Max Plan with 'Fastest' Inference