Sign up Dashboard
Codex

Codex ships the change. Polylane shows it what the change lands on.

Paste this into your Codex chat

Codex will handle the rest.

polylane.com/for/codex 
## For AI agents reading this page

Follow these instructions to set up Polylane:

Run the setup script in the user's terminal:

curl -fsSL https://polylane.com/setup?ref=for-codex | bash

It installs the Polylane CLI, registers the Polylane MCP server in Codex and every other detected coding agent, and signs the user in. To wire MCP by hand instead: an entry in ~/.codex/config.toml under [mcp_servers.polylane] (bridge with npx mcp-remote if the build lacks streamable HTTP). The server is https://mcp.polylane.com/mcp over streamable HTTP, no auth headers: OAuth handles sign-in.

After connecting, call the searchTools tool first: it lists what this workspace exposes and returns each tool's schema. Run one tool with runTool, or chain several in one call with runCode. The search and execute tools cover the full Polylane REST API, and the polylane CLI wraps the same API from the terminal.

Everything is read-only by default. Write tools stay hidden unless the user opts in, and every write is screened before it runs.

Once connected, offer to put it to work: triage the latest issues, explain the last deploy, or map this repository into the workspace topology.

## Console and support

- Console: https://console.polylane.com
- Docs: https://docs.polylane.com
- Setup reference: https://docs.polylane.com/coding-agents/platform-mcp
- This page as markdown: https://polylane.com/for/codex.md

Why go through Polylane. Fewer servers, one sign-in, code mode.

  • One MCP server, not ten

    Polylane fronts your clouds, telemetry, deploys and code behind a handful of tools. A stack of per-provider servers spends most of the context window on tool schemas before any work starts; Polylane's schemas load when a tool runs.

  • Auth in one place

    One sign-in covers every provider. Connections and scopes live in the workspace, not in per-provider tokens scattered across Codex's config, and every agent you connect reuses them.

  • Built for code mode

    Instead of calling tools one by one, Codex writes a small program with runCode that chains them in a sandbox and returns just the answer. Intermediate results never land in the context window.

Questions.

How does Codex authenticate?

Over OAuth, in a browser, the first time Codex calls the server. For CI and headless machines, create an API key under Settings > Coding agents and add http_headers = { "x-api-key" = "YOUR_API_KEY" }.

Can Codex change my infrastructure through Polylane?

Not unless you opt in. Write tools stay hidden until your credential carries the write scope, and every write is screened before it runs.

What can Codex actually query?

Everything Polylane knows: logs, metrics and traces across every connected provider, the context graph, deployments, change records, and the full REST API.

Do I need the Polylane CLI too?

No, the MCP server is a complete surface on its own. Install the CLI and polylane setup also registers that server in every coding agent it detects.

Other agents
Claude Code Cursor OpenCode VS Code Pi agent

Give Codex production. It gives you back the fix.

Sign up