WP-Autoplugin/wp-autoplugin

Quickly create functional plugins from simple descriptions, addressing specific needs without unnecessary bloat.

What it solves

WP-Autoplugin integrates an agentic coding workflow directly into the WordPress admin, allowing users to create, modify, and understand plugins and themes without leaving the site or manually moving code between an editor and the browser. It eliminates the risk of breaking live sites by staging changes in a separate workspace and requiring explicit administrator approval before release.

How it works

The tool uses a "Plan → Code → Review" pipeline powered by LLMs (via OpenAI, Anthropic, Google Gemini, xAI, or ChatGPT subscriptions).

  1. Plan: The agent analyzes the request and creates a versioned implementation plan and file map.
  2. Code: The agent generates code as immutable revisions. For existing projects, it uses bounded search-and-replace operations to avoid unnecessary rewrites.
  3. Review: An AI Reviewer evaluates the specific revision, providing a structured report with findings ranked by severity (P0-P3).

To avoid oversized prompts, the agent uses read-only source tools to progressively explore the codebase, inspecting metadata, searching text, and discovering WordPress hooks. Changes are staged separately and can be released as new plugins, forks, or direct modifications.

Who it’s for

WordPress developers and site administrators who want to automate the creation of plugins, fix bugs in existing themes/plugins, or extend third-party code via hooks without compromising site stability.

Highlights

  • Source-Aware Exploration: Progressively reads only relevant sections of the codebase rather than sending the entire project in every prompt.
  • Safe Staging: Code is never written to live files until an administrator approves the release.
  • Flexible Release Options: Supports building ZIPs, installing inactive plugins, creating forks, or applying direct modifications with rollback capabilities.
  • Vision Support: Allows users to upload screenshots or mockups to guide the AI in recreating interfaces or fixing layout issues.
  • BYOK (Bring Your Own Key): Compatible with multiple major AI providers and an experimental ChatGPT subscription flow.

Related

  • Project
    msrbuilds/elementor-mcp

    A WordPress plugin that exposes site management and page building tools to AI agents via the Model Context Protocol (MCP), enabling agents to build Elementor pages and manage site content.

  • Project
    WordPress/ai

    An experimental modular framework and plugin for WordPress that integrates AI-powered content creation, editorial tools, and site management features via external AI providers.

  • Project
    microsoft/skills-for-copilot-studio

    A plugin for Claude Code, GitHub Copilot CLI, and VS Code that enables authoring, testing, and troubleshooting Microsoft Copilot Studio agents using YAML files.

  • Project
    alibaba/open-code-review

    An AI-powered code review CLI tool that combines deterministic engineering with an agentic workflow to provide high-precision, token-efficient code analysis.

  • Dispatch
    Command Center: An Agentic Coding Environment for Production Quality

    Command Center is an agentic coding environment designed to reduce the time spent reviewing AI-generated code by providing structured walkthroughs and refactoring agents to ensure production-grade quality.