> ## Documentation Index
> Fetch the complete documentation index at: https://docs.enginy.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# ChatGPT & Codex

> Connect Enginy to ChatGPT — including Developer Mode custom connectors, now in beta on every paid plan — plus Codex for developers who prefer the CLI.

## ChatGPT

Most teams — especially non-technical roles like SDRs — will use Enginy from ChatGPT itself, not from a
terminal. This is the recommended starting point for OpenAI users.

<Info>
  On July 9, 2026, OpenAI folded its standalone Codex app into a single ChatGPT desktop app (macOS and
  Windows) with three modes — **Chat**, **Work**, and **Codex** — available on every plan, including Free. The
  previous ChatGPT desktop app was renamed **ChatGPT Classic**. The Codex CLI covered later on this page is a
  separate, independently-maintained tool from the Codex mode inside the new desktop app.
</Info>

### ChatGPT consumer app (Chat and Work)

The regular ChatGPT app — Chat and Work modes — only supports OpenAI-verified Apps from the official
directory. Enginy's verification is in progress, so Enginy isn't available there yet.

### ChatGPT Developer Mode

Developer Mode is the separate path that lets you add a custom, unverified MCP server like Enginy's. OpenAI
still labels it **beta**, and it's in beta across every paid tier — Plus, Pro, Business, Enterprise, and Edu —
though what you can *do* with it still depends on your plan:

* Full read/write access requires a **Business, Enterprise, or Edu** workspace, with an admin who explicitly
  enables write scopes. Individual **Plus/Pro** accounts stay **read-only** even with Developer Mode on.
* After an admin approves the connector, ChatGPT freezes the tool list at that point in time. Because Enginy's
  MCP surface is generated from the OpenAPI spec and can change (see [Overview](/mcp/overview)), tool calls
  can go stale until an admin reviews and republishes the connector.

To connect it:

1. **Personal Plus or Pro account**: go to **Settings → Security and login** and turn on **Developer mode**.
2. **Business, Enterprise, or Edu workspace**: a workspace admin goes to **Workspace Settings → Permissions &
   Roles** and enables **Developer mode** / custom MCP connectors for the org.
3. Either way, add `https://openapi.enginy.ai/mcp` as a custom connector and complete the browser approval
   flow.
4. Ask ChatGPT to use the Enginy connector and run `mcp_whoami` to confirm the connection and granted scopes.

## Codex

Prefer the terminal? The Codex CLI (or IDE extension) connects to the same hosted Enginy MCP server, and
isn't subject to the ChatGPT plan-tier gating above — it authenticates on its own.

### Before you start

An Enginy admin just needs to enable the MCP integration for the workspace — that's it to get started. Enginy
grants full access by default; restricting policy scopes is an optional, advanced step for later.

### Add the hosted server

Add Enginy with the Codex CLI:

```bash theme={null}
codex mcp add enginy --url https://openapi.enginy.ai/mcp
```

Then start the OAuth flow:

```bash theme={null}
codex mcp login enginy
```

Or configure it directly in `~/.codex/config.toml`:

```toml theme={null}
[mcp_servers.enginy]
url = "https://openapi.enginy.ai/mcp"
```

After the server is added, run `codex mcp login enginy` and complete the Enginy browser approval flow when prompted.

### Verify the connection

After authorization, ask Codex to use Enginy MCP and run:

```text theme={null}
mcp_whoami
```

That confirms the hosted MCP server is reachable, OAuth completed, and the granted scopes match what Enginy approved.

<CardGroup cols={2}>
  <Card title="Client support matrix" icon="table-list" href="/mcp/client-support-matrix">
    Compare support levels and setup shapes across every Enginy MCP client.
  </Card>

  <Card title="Security and troubleshooting" icon="triangle-exclamation" href="/mcp/security-troubleshooting">
    Debug auth, scope, token, and transport issues after ChatGPT or Codex connects.
  </Card>
</CardGroup>