Damon Cook, a long‑time WordPress developer and contributor now running his freelance shop Noma Digital, built WP Trend Watcher to solve a simple but painful problem: the relentless flood of WordPress news. Between core proposals, plugin and theme changes, community chatter, and event updates, staying informed takes hours each week. Damon wanted a smarter way to surface the important items without losing the human judgement that makes news useful.
What WP Trend Watcher does
WP Trend Watcher gathers a curated set of RSS feeds from WordPress sources, summarizes recent items with a local AI model, deduplicates and filters weak or repetitive pieces, and presents the results in a concise weekly briefing. Each briefing is a numbered list of headlines with short AI-generated summaries plus space for a human summary — Damon’s own commentary, verification notes and links — before publishing the report.
Why human review matters
Damon intentionally designed the tool with a “human in the loop.” The AI handles the heavy lifting of crawling, distilling and ranking content, but Damon reads through the suggested items, clicks through to validate them, and adds a short human summary. This dual workflow avoids pure automation pitfalls such as hallucinations or misleading headlines, and it helps the information actually register in a human brain. He publishes the time spent reviewing each report for transparency, so readers can judge the depth of review.
Who benefits
Freelancers, agency owners and developers get the most value. Knowing what’s coming in core releases and in developer discussions helps them anticipate client needs, adopt new features early, or spot breaking changes. For non-developers who still care about WordPress trends, it trims the noise and directs attention to what matters.
How it’s built (technical summary)
– Codebase: TypeScript, open source on GitHub so others can fork or contribute.
– Running cadence: Damon runs it weekly (configurable) and it captures new items since the last run.
– Local AI: He uses LM Studio on a Mac to run local models (choices vary by week). He experiments with different compact models that fit on his machine rather than relying on cloud APIs.
– Output: After Damon reviews and annotates, the report is published via GitHub Pages. The process is manual at the review step, so it’s not fully automated by design.
Filtering and fidelity
The system checks for duplicates (similar posts across multiple sites) and removes weak or superfluous items. Prompts are tuned to summarize specifically for developers and agency owners, which improves relevance. Damon treats the AI summary as a starting point: the fun part is fact‑checking it, clicking through, and adjusting the human summary when needed.
Philosophy and feelings about automation
Damon admits part of him feels “cheap” using AI to curate news, which is why he keeps the process transparent and public. He prefers not to fully automate publication because the human review both improves accuracy and makes the exercise useful to him personally. He enjoys the weekly ritual of reading, validating, and adding commentary, and he sees the project as scaffolding for richer trend analysis down the road.
Future possibilities
The project aims to be more than a weekly digest. By keeping past reports, there’s potential to use AI to analyze trends over time: why a feature proposal appeared, how discussion evolved, or where market shifts occurred. WP Trend Watcher is intentionally built so the approach could be applied to any vertical — not just WordPress — to build a searchable corpus and infer patterns across weeks, months and years.
How to try it
WP Trend Watcher is open source. Damon’s GitHub repository contains the code and example outputs (weekly summaries), and his blog explains how to get it running locally. It requires a machine capable of running local models (he runs it on a Mac with LM Studio) and a willingness to be the human reviewer for final publication.
Final thoughts
WP Trend Watcher sits in a useful middle ground: AI accelerates discovery and reduces dead time looking for links, while a human reviewer ensures accuracy, adds context, and keeps the output meaningful. For anyone overwhelmed by WordPress news — or curious how a local, human‑checked AI pipeline performs — it’s a practical, open approach worth exploring. Links to the repository, example reports, and Damon’s writeup are available in the show notes on WP Tavern if you want to dive deeper.