CLI

Connect OpenAI Codex CLI to Anoman AI

OpenAI Codex CLI routes through Anoman with a custom OpenAI-compatible provider in its config file — guardrails, routing, and cost tracking on every run.

Prerequisites

An Anoman API key — create one in the dashboard (format anm-sk-…).
OpenAI Codex CLI installed and up to date.
A model ID from the Anoman catalog, e.g. gpt-4o, gpt-4o-mini, or claude-sonnet-4-6.

Enable streaming for long agent runs

Set stream: true on requests. Autonomous, multi-step, and long-thinking tasks can exceed the 100-second edge timeout and fail with a 524 error; streaming returns tokens as they are generated and keeps the connection alive, so the request completes. Most tools stream by default — confirm it is on for agentic use.

More detail in the streaming guide.

Add Anoman as a model provider

Edit ~/.codex/config.toml (provider overrides only take effect at the user level). Pin wire_api = "chat" so Codex uses Anoman's /v1/chat/completions endpoint.

# ~/.codex/config.toml
[model_providers.anoman]
name = "Anoman"
base_url = "https://api.anoman.io/v1"
env_key = "ANOMAN_API_KEY"
wire_api = "chat"

model_provider = "anoman"
model = "gpt-4o"

Set your key and run

Export the key referenced by env_key, then start Codex — every request now flows through Anoman.

export ANOMAN_API_KEY=anm-sk-your-key-here
codex

Verify the guardrails are active

Every Anoman response carries guardrail result headers. Inspect them in your dashboard traces to confirm the pipeline is running on OpenAI Codex CLI requests — prompt-injection detection, PII masking, and content moderation all apply automatically.

X-Anoman-Guardrail-Injection: pass score=0.03
X-Anoman-Guardrail-Pii: pass
X-Anoman-Guardrail-Content: pass
X-Anoman-Cache: none

Need help with your integration?

Join the Founding 50 for early access and direct support from the Anoman team.

Join Founding 50