DEDUPED Hot AI repos, turned into use cases Get the Monday 5

How to point Codex CLI at the gateway with a provider entry with Codex LB Enhanced

Add a Codex LB Enhanced provider to config.toml so Codex routes through the pool.

Credentials. The config file will hold a reference to an API key, so keep the file private and do not commit it to version control.

Who it is forA developer who uses the Codex CLI or IDE integration and wants long sessions to survive account quota limits and interruptions.

The problemCodex sessions break when one upstream account is exhausted or a WebSocket drops without a close frame.

The outcomeCodex sends its Responses traffic through the gateway, with WebSocket support enabled and the provider selected by default.

What you need first

  • Docker installed and running, or uv installed for running from source
  • Upstream ChatGPT accounts you are authorized to use, to add in the dashboard
  • A browser to open the dashboard at localhost:2455
  • Codex CLI, desktop, or IDE integration, or another OpenAI-compatible client, if you plan to connect one

Steps

  1. Start the gateway and create an API key in the dashboard at http://localhost:2455.
  2. Open ~/.codex/config.toml and add the model, model_reasoning_effort, and model_provider lines from the README example.
  3. Add a [model_providers.codex-lb-enhanced] block with base_url set to http://127.0.0.1:2455/backend-api/codex, wire_api = "responses", supports_websockets = true, and requires_openai_auth = true.
  4. Replace base_url with your reverse-proxy address if you do not run the gateway on localhost.
  5. Start Codex and send one short prompt to confirm it reaches the gateway.

Prompt for your agent

Using the aafqaq/codex-lb-enhanced README, add the codex-lb-enhanced provider block to <path to config.toml> with the base_url for <gateway URL>. Report the final provider block and whether a test Codex prompt succeeds. Verify that the file still parses as TOML.

Caveat. The README example uses a specific model name, so change model and model_reasoning_effort to values your account supports.

Use this in your AI agent

Ask your agent

Using the Deduped MCP, get the setup steps for aafqaq/codex-lb-enhanced and "Point Codex CLI at the gateway with a provider entry"

Add Deduped to Claude Code once

claude mcp add --transport http deduped https://dedupedai.com/mcp

Cursor, ChatGPT, others

Share: X LinkedIn Hacker News Email Copy link

Third-party repos we don't own. AI-written use cases can be wrong. Disclaimer