Codex Knowledge Base
Codex: an agentic software engineering knowledge base
A continuously updated knowledge base for Codex, covering agentic software engineering workflows, configuration and real-world integration. New articles publish as the tool evolves.
This knowledge base complements the book Codex CLI: Agentic Engineering from First Principles. The map above is the book's framework — every article here goes deeper on a piece of it.
Get the bookThe 2× Rework Tax: Late-Arriving Requirements and Code Invalidation in Coding-Agent Sessions
The 2× Rework Tax: Late-Arriving Requirements and Code Invalidation in Coding-Agent Sessions
Harness Engineering: Anatomy, Architecture, and the Seven Subsystems Underpinning Codex CLI
Harness Engineering: Anatomy, Architecture, and the Seven Subsystems Underpinning Codex CLI
Front-Load or Fail: The Four-Phase Coding Agent Workflow and Its Codex CLI Implementation
Front-Load or Fail: The Four-Phase Coding Agent Workflow and Its Codex CLI Implementation
Codex CLI v0.153.0: Three-Tier Context Management, Experimental Token-Budget Mode, and Guardian History Fencing
Codex CLI v0.153.0: Three-Tier Context Management, Experimental Token-Budget Mode, and Guardian History Fencing
Agentic Cloud Engineering: How Graph, Loop, and Zero-Trust Harness Abstractions Make Bounded Autonomous Work Provable
Agentic Cloud Engineering: How Graph, Loop, and Zero-Trust Harness Abstractions Make Bounded Autonomous Work Provable
Working Memory Is Not a Token Count: Coherence Debt, Heterogeneous Memory Objects, and Codex CLI's Experimental Context Management Mode
Working Memory Is Not a Token Count: Coherence Debt, Heterogeneous Memory Objects, and Codex CLI’s Experimental Context Management Mode
Don't Let the Model Write the YAML: Deterministic GitOps Remediation and the Codex CLI Configuration Safety Problem
Don’t Let the Model Write the YAML: Deterministic GitOps Remediation and the Codex CLI Configuration Safety Problem
Refusing the Impossible: What the Code Hallucination Benchmark Reveals — and How Codex CLI Hooks Close the Gap
Refusing the Impossible: What the Code Hallucination Benchmark Reveals — and How Codex CLI Hooks Close the Gap
Asynchronous TUI Questions in Codex CLI v0.154: A New Paradigm for Human-Agent Collaboration
Asynchronous TUI Questions in Codex CLI v0.154: A New Paradigm for Human-Agent Collaboration
Non-Atomic Tool Failures: Verified Tool Calls, Idempotency Keys, and the 52pp Duplicate-Reduction Pattern for Codex CLI Agents
Non-Atomic Tool Failures: Verified Tool Calls, Idempotency Keys, and the 52pp Duplicate-Reduction Pattern for Codex CLI Agents