How to use Unknown to automate blog SEO
The author reports that saved drafts now receive SEO metadata, summaries, and categorization automatically, with Next.js rebuilding pages using the generated metadata.
dev.toPrompt to try
Editorial adaptation, not the source author’s exact prompt. Replace bracketed inputs and review tool permissions first.
Analyze [Markdown draft] as an SEO specialist and content editor. Return JSON containing a short summary, difficulty level, content intent, Open Graph title, primary keyword, and secondary keywords. Base every field on the supplied draft.
Reach people exploring AI agents
See placements →What happened
The author reports that saved drafts now receive SEO metadata, summaries, and categorization automatically, with Next.js rebuilding pages using the generated metadata.
Reported by yabasha on dev.toHow the agent did it
- Write a rough technical article in Markdown.
- Save the draft so PostObserver queues analysis.
- Run the editor agent to generate metadata.
Where a person stepped in
The author writes the draft and chooses the active model; the post does not describe reviewing generated metadata.
A person set the goal and reviewed the result.
The author built a custom editor agent for Yabasha.dev to reduce the administrative work surrounding technical articles. Laravel 12 handles the backend and a Filament admin panel, while Next.js 15 serves the frontend. OpenRouter supplies access to different models without tying the service to one provider.
The dedicated PostAiService constructs a prompt asking for an SEO specialist and content editor. Its structured output includes a short summary, difficulty level, content intent, Open Graph title, primary keyword, and supporting keywords. The author also created a Filament model resource and a benchmarking method, allowing models to be sorted by cost and measured speed before changing the active model.
The publishing sequence starts with a rough Markdown draft. Saving it triggers a PostObserver, which queues the analysis job. The agent reads the draft and fills in metadata, summaries, and categorization. Next.js then rebuilds the page and incorporates that metadata into the document head. The author reports that this removed the blogging administration that had discouraged publishing. Who it's for: developers maintaining a technical blog who want draft analysis integrated into their publishing stack.
Prompt
Suggested prompt (written by Agents In Use, not the author's)
Source: dev.to · yabasha
Frequently asked questions
What starts the metadata generation?
Saving a Markdown draft causes PostObserver to detect the change and queue a job for the editor agent.
Can the author change the model?
Yes. A Filament resource tracks OpenRouter models, and the admin panel lets the author sort by cost and speed and toggle the active model.
More details
Remaining steps
- Rebuild the Next.js page with the metadata.
About this account
The background agent analyzes saved drafts and fills structured publishing metadata after a queued job starts.
Platform: dev.to. Attribution: https://dev.to/yabasha/how-i-built-an-ai-agent-for-my-portfolio-yabashadev-using-laravel-nextjs-52a9. Prompt: Suggested (written by Agents In Use, not the author's). Sourced by the Agents In Use pipeline.
Credit: yabasha on dev.to
Source published 19 Dec 2025
Source checked 14 Oct 2026
Tools and topics
Laravel 12 · Next.js 15 · Filament · OpenRouter
Blogging · SEO · Metadata
Editorial completeness: 100/100. Measures documented fields, not agent performance or verified outcomes.