LessAI ToolsAI plugin directory

How to add MCP to Codex: start with a documentation server

Understand local and remote MCP connections in Codex. Add the official read-only OpenAI documentation service, check its configuration and verify an actual call.

Start with a working Codex CLI

Install and sign in to Codex CLI using its official guide. CLI means the command-line tool run in a terminal; this guide uses it to configure MCP. Start with available Codex access rather than purchasing a new plugin service.

MCP supplies external tools and information. Adding a server does not retrain the model or automatically grant every permission.

Choose a local command or a remote address

A local STDIO server starts through a command; check its runtime and arguments. A remote Streamable HTTP server uses an address and may require authorization.

The default configuration is ~/.codex/config.toml; trusted projects can use .codex/config.toml. Clients on the same Codex host share configuration. Another device or cloud environment still needs its own connectivity checks.

Try the official read-only documentation example

OpenAI’s public documentation MCP reads developer documentation; it does not call the OpenAI API for you. These are official setup examples, not commands executed by this website.

  1. In a terminal with Codex CLI installed, run: codex mcp add openaiDeveloperDocs --url https://developers.openai.com/mcp
  2. Run: codex mcp list. Check for openaiDeveloperDocs.
  3. Use /mcp in the Codex terminal interface to inspect the active connection, then ask a documentation question with sources.

Saved configuration is not a successful connection

A listed entry confirms configuration, not a working call. Check startup, authorization and an actual result. For an OAuth server, follow the provider’s requirements and codex mcp login server-name; not every server needs login.

  1. For a local failure, check the runtime, executable and arguments.
  2. For a remote failure, check the address, network and required authorization.
  3. After changes, restart or reconnect as your client instructs and verify a call.

Add tools for a specific task

Choose a service for the task, read its source, permissions and setup, then try a checkable request. Keep credentials in appropriate local configuration, outside public repositories.

Sources checked on 2026-10-12. Installation, login and connectivity were not tested here. Follow current client and server guidance.

Sources and ranking

This directory lists AI development plugins with public source code and a clear purpose. Browse by task, then open an introduction to learn what a project does, who it is for, and what you need to get started. It covers selected projects, not every AI plugin.

How do the three plugin types differ?

Tool connections (MCP) let AI assistants use external tools such as browsers, code repositories and databases. MCP is the protocol used for these connections.

Reusable skills give a compatible AI assistant instructions and guidance for a particular task.

Editor plugins run inside the software you use to write code, adding AI chat, completions or suggested edits.

How does “Most starred” work?

Projects are sorted by their total GitHub stars. Stars indicate interest; they are not downloads, active users or quality scores. A skill pack counts as one project.

What does “Recently updated” mean?

It uses the most recent code push, which may not be a new release. The 7-day filter checks that project update time; ranking still uses all-time stars. Each introduction shows when its star count was collected.

Introductions are based on public project documentation. These projects have not each been installed and tested here. Check the project documentation for setup, compatibility and service costs.

UTC+8

Pass a useful find along

Pick a draft and a platform. You can edit before posting.