Quick Answer
I built mohdanish.me with Nuxt (SSR), Supabase for content, and a small admin CMS so I can publish without redeploying for every edit. The goal was a fast, crawlable portfolio that also hosts long-form technical posts — not a static brochure that dies after launch day.
Quick Facts
| Item | Details |
|---|---|
| Topic | Nuxt portfolio + blog architecture |
| Category | Nuxt / Supabase / SEO |
What Is This Stack?
This is not “drop a theme on Vercel and call it a portfolio.” The site is a Nuxt app with server-side rendering so Google gets real HTML on first request. Content (projects, about copy, blog posts) lives in Supabase. An admin area lets me edit and publish without shipping a new deploy for every typo.
If you only need a one-page resume PDF online, this is overkill. I needed a place that can grow into a blog people actually land on from search — Laravel, Nuxt, servers, auth — while still showing client work.
Why It Matters
- Crawlable content — SSR + sitemap + canonicals beat a client-only SPA for article indexing.
- Editable without fear — blog posts and project text update from admin, not from a git commit every time.
- One domain for two jobs — portfolio proof + technical writing, so inbound traffic is not stuck on a lonely Contact form.
How It Works
Three layers stay separate on purpose.
- Nuxt front — pages, layouts, SEO meta, JSON-LD for BlogPosting / FAQ / HowTo when the HTML earns it.
- Supabase — rows for site content and
blog_posts; RLS keeps public reads open and admin writes gated by auth. - Admin CMS — rich text for structured posts, image CDN for covers, soft-delete and view counts when I need them.
Publish flow: write in admin → status published → sitemap refresh route warms /sitemap.xml so new URLs show up without waiting for a full redeploy.
Step-by-Step Guide
Step 1: Lock SSR and site URL
Keep Nuxt in SSR mode. Set a single canonical site URL in config (no trailing-slash fights). Done check: View Source on / and a blog slug — you should see the article title in the raw HTML, not an empty app shell.
Step 2: Put content in Supabase, not hard-coded arrays
Move projects, about paragraphs, and posts into tables. Public pages read with the anon key under RLS. Done check: change a project description in admin, refresh the home page, see the new text without rebuilding.
Step 3: Ship sitemap + robots before you write ten posts
Wire a dynamic sitemap to published slugs. Add robots.txt that points at the sitemap and blocks /admin. Done check: open /sitemap.xml and confirm every published post URL is listed with a sensible lastmod.
Real-World Example
Stack on day one: Nuxt + Supabase Auth for admin + public blog table. Goal: publish a Laravel Sanctum post and have it indexable the same week.
Failure mode: I published the post, but the sitemap was still cached or generated from an empty list, so Search Console sat on “Discovered – currently not indexed.” Soft 404 risk is real if a slug fetch fails and you return an empty shell with 200.
Fix: disable aggressive CDN cache on /sitemap.xml, add an admin “refresh sitemap” call after publish, and make single-post fetch errors return 503 (temporary) instead of a fake 404. After that, new posts appeared in the sitemap within minutes.
Pros & Cons
Advantages
- Real HTML for crawlers without maintaining a separate static export pipeline.
- Blog and portfolio share one design system — fewer “this page looks abandoned” vibes.
- I can iterate on SEO fields (meta title, OG image, tags) per post from admin.
Disadvantages
- More moving parts than a pure static site (auth, RLS, env keys, image CDN).
- If Supabase is down, public content fetch needs honest error pages — not blank marketing copy.
- You still have to write the articles. The stack does not invent expertise.
Best Practices
- Write posts from real bugs (419 CSRF, CORS credentials, hydration mismatches) — that is the unique value.
- Keep admin noindex; never leak draft HTML to public routes.
- Refresh the sitemap on publish; do not wait for the next full deploy.
- Moderate comments before they go live if you allow UGC next to ads later.
- Prefer one solid 1,200-word guide over five thin stubs.
Common Mistakes
- Client-only portfolio — why it happens: faster to ship a SPA. Fix: SSR (or prerender) any page you want indexed.
- Hard-coded blog markdown in the repo only — why it happens: feels simpler. Fix: fine for a few posts; painful once you edit from your phone.
- Returning 200 with empty content on DB failure — why it happens: catch-all UI. Fix: 503 for outages, 404 only when the slug truly does not exist.
- Publishing a 150-word launch announcement forever — why it happens: placeholder left live. Fix: expand it into a real guide or unpublish it.
Frequently Asked Questions
What stack did you use for this portfolio?
Nuxt for the app and SSR, Supabase for content and admin auth, plus standard SEO plumbing (meta tags, JSON-LD, sitemap). Images go through a CDN workflow so covers stay fast.
How do you update blog posts without redeploying?
Posts live in Supabase. The admin editor writes HTML + SEO fields; public pages read published rows. A sitemap refresh after publish keeps Google’s list in sync.
Nuxt vs a static site generator for portfolios?
Static is enough for a resume page. Nuxt SSR helped once the blog and admin CMS became first-class — especially for per-request meta and avoiding empty shells on article URLs.
Is this setup worth it for freelancers?
Yes if you plan to publish technical writing on the same domain as your work. If you never write, a simpler static site is less to maintain.
Summary
The portfolio is a Nuxt + Supabase content site: SSR for crawlers, admin for edits, sitemap discipline so new posts are discoverable. The stack only pays off if the writing is real and maintained.
Key Takeaways
- SSR + honest status codes matter more than fancy animations for indexing.
- Put blog content in a CMS/database if you expect frequent edits.
- Refresh the sitemap when you publish — do not assume deploys will catch up.
- Thin launch posts hurt trust; expand them or take them down.
- Unique value comes from lived debugging notes, not from the framework logo row.

Comments
0 comments · new ones appear after approval
No comments yet. Be the first to share your thoughts.
Leave a comment
Your comment will be reviewed before it appears.