wtf is seo-decision-lab?

SEO Decision Lab is a plugin for a tool called Codex, built to support SEO strategy and day to day SEO work. It connects to a service called DataForSEO and adds a local server, using the MCP protocol, plus a skill you can call inside Codex to do keyword research, check competitors, look at search result pages, run on page audits, check backlinks, look at local SEO signals, monitor content and reputation, and track how a brand shows up in AI generated search answers. To install it, you add the repository to Codex as a plugin marketplace and then install the seo-decision-lab plugin from there. The plugin itself lives in a folder called plugins/seo-decision-lab, and if you want to test it outside of Codex you can go into that folder and run standard npm install and npm run smoke commands. This is a bring your own key tool, meaning every user has to supply their own DataForSEO account credentials after installing. Inside a Codex conversation, you can simply ask it to check the DataForSEO authentication status, and if credentials are missing Codex will prompt you to save them. These are stored locally on your own machine, not sent anywhere else, in a config file under your home directory. You can also set them through environment variables instead of the interactive prompt. Once configured, the plugin exposes a set of tools that Codex can call directly, covering authentication status, saving and clearing credentials, checking your DataForSEO account status, listing available endpoints, making raw requests, running live Google search result checks, generating keyword ideas, getting keyword overviews, viewing ranked keywords, summarizing backlinks, searching for mentions in AI generated answers, and posting or checking on page audit tasks. The project explicitly avoids shipping any API keys, and its README asks contributors not to commit local credential files.env files, or Codex configuration files that might contain secrets. This is a specialized tool for people already using Codex who want SEO research built into their workflow, rather than a general purpose product for a wide audience.