🧰 OpenAI Codex

The items that name it 31

• [tl;dr sec] #352 - Unprompted Slides, AWS AI Post-compromise Trends, Malicious LLM API Routers
  articleas affected products2026-10-29It is written here as OpenAI Codex (sessions exposed through poisoned routers).

  sessions exposed through poisoned routers

• Buchi Reddy B (@buchireddy) on X
  threadas affected products2026-10-27It is written here as OpenAI Codex (coding agent): wrote an attacker's MCP server into its machine-wide config during the Forever Security test.

  coding agent

• Microsoft Targets Rogue AI Agents With Dynamic Permissions, Risk Scoring
  articleas affected products2026-10-20It is written here as OpenAI Codex (local agent, among 60 supported types for discovery).

  local agent, among 60 supported types for discovery

• Thinking in Graphs: Neo4j-Backed Exploit-Chain Reasoning for LLM Penetration-Testing Agents
  paperas tools2026-10-14It is written here as OpenAI Codex (LLM coding agent that can drive penetration-testing tools; reproducible setup given).

  LLM coding agent that can drive penetration-testing tools; reproducible setup given

• Zero to calc in six days: inside XBOW's Chrome full chain
  writeupas classification.technologies2026-10-13
• Codex CLI installs a persistent, self-updating background daemon by default
  threadas classification.technologies2026-10-12
• Google Pixel 10 Exploits Earned Hackers $560,000 at Pwn2Own
  articleas affected products2026-10-09

  Coding tool exploit earned $40,000

• Vikram Dias (@BigVikDada) on X
  threadas classification.technologies2026-09-29
• OpenAI DevDay 2026 live blog
  writeupas classification.technologies2026-09-29
• AI Agent Hacking Writeups
  repoas affected products2026-09-17
• Codex with GPT-6 Astra posts 82.1% FuncPass and 34.1% SecPass | Blog | Endor Labs
  writeupas classification.technologies2026-09-08
• GitSpawn: A Single Flaw Lets Untrusted Repos Run Code in Claude Code, Codex, Cursor, and Grok - Manifold Security
  writeupas affected products2026-09-01

  affected by a variant of the same flaw class (different mechanism); tested and reported, closed as duplicate of another researcher's report; since patched

• The AI Attack Surface: How Threat Actors Abuse Trusted AI Platforms
  writeupas affected products2026-08-27

  Impersonated in a later variant reported by Cato Networks involving a fake download

• MinIO Security Lab
  repoas tools2026-08-21It is written here as OpenAI/codex.

  cross-vendor independent code-review verifier used in the analysis pipeline

• Recovering Encrypted LLM Reasoning Traces
  writeupas affected products2026-08-17It is written here as OpenAI Codex (session files under ~/.codex/sessions/2026/08/14 contain encrypted reasoning payloads).

  session files under ~/.codex/sessions/2026/08/14 contain encrypted reasoning payloads

• 0DIN Expands: From Inference to Agency
  articleas affected products2026-08-15

  in-scope agentic product

• Codex Security CLI
  repoas classification.technologies2026-08-01
• AI Security Briefing
  advisoryas classification.technologies2026-07-18
• AI Security Briefing
  articleas affected products2026-07-18

  One of the AI models that can power GitHub Agentic Workflows

• Skills Janitor
  repoas affected products2026-07-15

  Also supported by skills-janitor for skill scanning

• Introducing Patch the Planet
  articleas classification.technologies2026-06-22
• One Fake Bug Report Hijacked a $250B Company’s AI Agent
  articleas affected products2026-06-17

  AI coding agent; sandboxed instance in CI/CD (CircleCI on EC2, CODEX_SANDBOX_NETWORK_DISABLED) and as VS Code extension (ChatGPT/Codex) on macOS, both reached.

• The sorry state of skill distribution
  writeupas affected products2026-06-03

  Agent harness that can accept skill ZIP archives uploaded manually or via API

• Legitimate-Looking Codex Remote UI Secretly Steals Your AI Tokens
  writeupas classification.technologies2026-05-27
• KeygraphHQ/shannon
  repoas classification.technologies2026-05-23
• GenAI App Security Checklist
  repoas classification.technologies2026-05-13
• Mythos AI Security Testing
  articleas classification.technologies2026-05-12
• Pantheon-Security/medusa
  repoas affected products2026-05-10It is written here as OpenAI Codex (AGENTS.md).

  AGENTS.md

• Untangling a Linux Incident With an OpenAI Twist (Part 2)
  writeupas tools2026-04-22

  AI coding agent used by legitimate developer to build applications and (ineffectively) troubleshoot/respond to malicious behavior on the Linux system; masked symptoms like fan noise via CPU throttling but lacked root access, continuous monitoring, and threat intel context

• Untangling a Linux Incident With an OpenAI Twist
  writeupas tools2026-04-17

  AI coding agent used by the legitimate end user to develop two web applications and to attempt troubleshooting/security audits and remediation of suspected malicious activity (e.g., CPU throttling suggestion, killing a cryptominer instance); its generated shell commands were also picked up by EDR de

• Plugin4Shell - Zero Click RCE Vulnerability found in top 4 most popular coding agents, millions of agents affected
  writeupas affected products

  Affected by Plugin4Shell; patched after disclosure in version 0.146.0

In every workspace you can read 37