OpenAI Codex CLI
Codex CLI is OpenAI's open-source agentic coding tool. It connects to AgentXchain via the local_cli adapter.
Which adapter?
local_cli — AgentXchain spawns codex as a subprocess for each governed turn.
Prerequisites
- Codex CLI installed (
npm install -g @openai/codex) OPENAI_API_KEYset in your environmentagentxchainCLI installed
Configuration
{
"runtimes": {
"codex-dev": {
"type": "local_cli",
"command": ["codex", "--quiet", "--full-auto"],
"cwd": ".",
"prompt_transport": "argv"
}
},
"roles": {
"dev": {
"runtime": "codex-dev",
"mandate": "Implement features and fix bugs",
"authority": "proposed"
}
}
}
Key fields
| Field | Value | Why |
|---|---|---|
command | ["codex", "--quiet", "--full-auto"] | --quiet suppresses interactive output. --full-auto allows file writes without confirmation. |
prompt_transport | "argv" | The prompt is passed as a command-line argument. Use "stdin" if your Codex version supports piped input. |
Verify the connection
agentxchain connector check
Minimal working example
mkdir my-project && cd my-project
agentxchain init --governed --template cli-tool --goal "Build a JSON schema validator" -y
git init && git add -A && git commit -m "initial scaffold"
agentxchain run
Gotchas
--full-autois required for non-interactive governed execution. Without it, Codex will prompt for approval on file changes.- Model selection: Codex CLI defaults to a Codex-optimized model. You can override with
--model gpt-5.4if needed, but the default is usually better for coding tasks. - Working directory: Codex CLI operates on the current working directory. The
cwdfield ensures it runs in your project root.