About This Project

This plugin is my application project for the Fusion AI Coders Program. The goal is to demonstrate problem-solving skills, effective use of AI tools, and ability to work with WordPress development.

Initial Setup & Ideation

AI Configuration

I use an ai-config-hub, a personal tool I built to centralize AI configurations across my projects. It manages Claude Code settings, MCP servers, and other AI tool configurations through a single command (ai-init), making it easy to maintain consistent AI workflows across different projects. This includes reusable slash commands, specialized agents, and a base CLAUDE.md configuration—essentially a Claude Code boilerplate for repeated tasks and coding patterns I use across all my projects.

Research & Brainstorming

To explore the project options, I used the context-a8c MCP plugin to fetch and read the AI Coders Program application post directly within Claude Code. This allowed me to:

1. Review all five project options without leaving my development environment
2. Brainstorm ideas with Claude Code for each option (Dashboard Widget, Settings Page, Admin Column, Content Add-On, Custom Block)
3. Evaluate trade-offs between complexity, visual impact, and practical utility

After discussing several ideas—including a "Content Freshness Dashboard" and "Post Publishing Checklist"—we landed on the Post Health Score concept for Option 3 (Admin Column). This choice balanced sophistication with simplicity, creating something genuinely useful for content managers while demonstrating solid WordPress development skills.

Post Health Score

A WordPress plugin that adds a content quality "Health Score" column to the Posts admin list.

Option Chosen: Admin Column

I chose Option 3: Admin Column from the available project options because:

1. Practical Utility: A content health score is genuinely useful for content managers and editors to quickly identify posts that need improvement.

2. Technical Depth: This option allows me to demonstrate understanding of WordPress hooks, post meta, taxonomies, and admin customization - all without being overly complex.

3. Visual Impact: The result is immediately visible and testable in the WordPress admin, making it easy to verify functionality.

4. Creativity Opportunity: The scoring system and visual presentation (grades, emojis, tooltips) allow for creative expression while solving a real problem.

Plugin Goal

Add a "Health Score" column to the Posts list in WP Admin that grades each post based on content quality indicators:

Criterion	Requirement	Points
Word Count	> 300 words	+1
Featured Image	Has one set	+1
Title Length	30-60 characters	+1
Categories	At least 1 (excl. "Uncategorized")	+1
Tags	At least 1	+1

Grades:

• 5 points = A+ (Excellent)
• 4 points = A (Great)
• 3 points = B (Good)
• 2 points = C (Needs Work)
• 1 point = D (Poor)
• 0 points = F (Failing)

Planned Features

• [x] Health Score column in Posts list
• [x] Scoring based on 5 content quality criteria
• [x] Letter grade with emoji indicator
• [x] Hover tooltip showing detailed breakdown
• [x] Color-coded badges for quick scanning
• [x] Sortable column

Implementation Plan

Phase 1: MVP

1. Create minimal plugin structure
2. Add column to Posts list
3. Implement basic scoring logic

Phase 2: Enhancement

4. Add grade letters and emoji display
5. Add CSS styling for badges
6. Add tooltip with detailed breakdown

Phase 3: Polish

7. Make column sortable
8. Final documentation and cleanup

AI Tools Used

• Claude Code with Claude Opus 4.5 - Primary development assistant for planning, code generation, and problem-solving
• Codex - Code review and quality assurance (identified 5 issues that were subsequently fixed)

Installation

1. Download or clone this repository
2. Upload to your WordPress site:
   • Option A: Unzip and upload the folder to wp-content/plugins/
   • Option B: Upload the .zip file directly via Plugins > Add New > Upload Plugin in wp-admin
3. Activate the plugin in WordPress admin

Development Process

The plugin was developed incrementally, following a phased approach:

What Worked Well

1. Incremental Development: Building the plugin step-by-step made it easier to test each feature in isolation and catch issues early.

2. Claude Code Assistance: The AI helped with:

   • Planning the implementation approach
   • Suggesting WordPress hooks and functions
   • Writing clean, documented PHP code
   • Generating CSS for tooltips and badges
   • Reviewing code structure

3. Starting with README: Writing the README first (before any code) helped clarify the project scope and requirements.

4. MVP First: Getting a basic working column before adding visual enhancements allowed for faster iteration.

What I Learned

• WordPress admin column hooks (manage_posts_columns, manage_posts_custom_column)
• How to make columns sortable with manage_edit-post_sortable_columns and pre_get_posts
• Using post meta to cache calculated values for efficient database queries
• CSS-only tooltips as an alternative to JavaScript-based solutions

Challenges & Solutions

Challenge 1: Column Sorting

Problem: Health scores are calculated dynamically, but WordPress needs a database field to sort by.

Solution: Store the score as post meta (_phs_health_score) when the column is rendered and when posts are saved. This provides a cached value for efficient sorting.

Challenge 2: "Uncategorized" Category

Problem: Every post has at least the "Uncategorized" category by default, which would make the category check always pass.

Solution: Explicitly exclude category ID 1 (Uncategorized) from the count when checking for meaningful categories. Updated: Use get_option('default_category') to dynamically get the actual default category ID, since it's configurable and may not always be ID 1.

Challenge 3: Tooltip Positioning

Problem: Tooltips need to appear above the badge without being clipped by the table row.

Solution: Used CSS positioning with position: absolute, bottom: 100%, and a high z-index to ensure visibility.

Code Review & Quality Assurance

Following the project tip to "ask another model to review your code", I used Codex to review the initial implementation. This cross-AI review identified 5 issues:

Issues Found & Fixed

Issue	Severity	Problem	Fix
Hard-coded category ID	Medium	Assumed ID 1 is always "Uncategorized"	Use get_option('default_category')
Unnecessary DB writes	Medium	update_post_meta ran on every page view	Removed from render, kept only on save_post
ASCII-only functions	Medium	strlen() and str_word_count() fail for non-Latin text	Use mb_strlen() and preg_split()
Non-translatable strings	Low	Labels weren't wrapped in __()	Added load_plugin_textdomain() and wrapped strings
CSS scope too broad	Low	Styles loaded on all post types	Added $screen->post_type check

Lessons Learned

1. Cross-AI review is valuable: Having a different AI model review the code caught issues that weren't obvious during initial development.

2. Internationalization matters: Even for a demo project, proper i18n shows attention to WordPress standards.

3. Performance considerations: Avoiding unnecessary database writes on read operations is important for scalability.

Screenshots

After activating the plugin, you'll see a "Health Score" column in the Posts list:

Low Score	High Score
Hover shows which criteria failed	Hover shows all criteria passed

File Structure

post-health-score/
├── post-health-score.php    # Main plugin file (~310 lines)
├── assets/
│   └── css/
│       └── admin.css        # Admin styling (~107 lines)
└── README.md                # This file

License

GPL v2 or later

---

Built for the Fusion AI Coders Program application.