OpenAI - Unrolling the Codex agent loop

Evidence record

Read the original source ↗

Claim-level citations (1)

• supportsContext privilege escalation let repository instructions blind an AI code reviewer: OpenAI's Codex documentation confirms that repository AGENTS/AGENTS.override instructions are aggregated into the model context with more-specific instructions taking precedence, establishing the relevant context-assembly behavior independently of the attack paper.

  Constructing the initial prompt, repository AGENTS.md and AGENTS.override.md instructions