Codex as a platform: build on the open agent harness | OpenAI Developers
Aug 19, 2026

Summary
This post introduces the open-source Codex harness as a foundational component for building agentic applications. It details how the harness provides a reusable agent loop for context management, tool use, sandbox enforcement, and approval workflows. The post emphasizes integrating this harness into custom applications via the Codex app-server or SDK, allowing developers to build software around specific workflows (e.g., security investigations, operations dashboards) rather than reproducing a generic chat interface. It highlights the flexibility of exposing application-owned context, tools, and approval flows to the agent, exemplified by the Relay operations application.
More from the Agentic Execution Environment trail
- Bringing my LED display to life with GPT-Live-1 and Codex | OpenAI Developers
- Rethinking skills and prompts for GPT-6 Astra | OpenAI Developers
- Meet Rosalind Workbench: Empowering every scientist to be their own research team | OpenAI Developers
- Automating repetitive work at OpenAI with Codex
- Custom Code Review rules for Codex | OpenAI Developers