Worked examples bolt.new
Is bolt.new vibe coded?
StackBlitz's answer to the same question. We fetched the homepage on 29 August 2026 and ran all 86 detectors over the raw HTML — the same scan you get from the box on the front page, frozen here so the numbers can be argued with.
bolt.new
Hand-written
Somebody sat down and made decisions. Several of them were opinions.
- Vibe tells
- 6
- Craft tells
- 5
- Confidence
- high
Plenty of signal to work with.
Where the score came from
- Platform fingerprints1 hit
Artefacts of the tool, not of taste
- Colour, type and surface2 hits
The visual defaults, applied unedited
- Layout and architecture3 hits
The shape a page takes when nobody asks what it needs to say
- Copy tellsnone
How the words were written
- Leftoversnone
Scaffolding never cleaned up
- Source and buildnone
What is visible in the shipped code
- How it behavesnone
Weight, mobile, accessibility, SEO — where the money leaks
- Human craft5 hits
Evidence of deliberate decisions
Evidence
11 findings- +18
Bolt / StackBlitz fingerprints
Bolt or StackBlitz artifacts are shipping with the page.
- +9
The typeface nobody chose
None of these are bad faces. The tell is that they arrive as the default — the page reads as though nobody made a typography decision.
- inter
- −9
Deliberate social preview
Open Graph title and image were both authored.
- +8
Exactly three feature cards in a row
Features arrive in threes, as tradition demands — regardless of how many things the product actually does.
- +7
Everything centred
18 centred blocks. Long centred text is measurably harder to read, and with no left-aligned anchor the hierarchy flattens.
- −7
Hand-written ARIA
14 explicit accessibility attributes.
What we read
- title
- Bolt AI builder: Websites, apps & prototypes
- generator
- —
- server
- cloudflare
- status
- 200
- html
- 115.3 KB
- words
- 504
- fetched in
- 710 ms
Reading this score
bolt.new landed on 15/100, which puts it in the hand-written band. That number is not a percentage and not a probability — it is strength of evidence: 32 points of tells, damped by craft, less 32 points of deliberate decisions, mapped through one logistic curve.
What pushed it up
- Bolt / StackBlitz fingerprints +18 — Bolt or StackBlitz artifacts are shipping with the page.
- The typeface nobody chose +9 — None of these are bad faces. The tell is that they arrive as the default — the page reads as though nobody made a typography decision.
- Exactly three feature cards in a row +8 — Features arrive in threes, as tradition demands — regardless of how many things the product actually does.
- Everything centred +7 — 18 centred blocks. Long centred text is measurably harder to read, and with no left-aligned anchor the hierarchy flattens.
- Spacing that doesn't keep time +5 — 6 different vertical rhythms across 7 sections, because each one was generated in its own prompt. The page reads as unfinished even to people who can't say why.
- A bento grid because bento grids exist +5 — Asymmetric tiles used as a style rather than because some tiles matter more. If shuffling them changes nothing, the grid is decoration.
What pulled it down
- Deliberate social preview −9 — Open Graph title and image were both authored.
- Hand-written ARIA −7 — 14 explicit accessibility attributes.
- Structured data −6 — JSON-LD is present — somebody thought about search.
- Semantic landmarks −6 — 4 landmark elements instead of a pile of divs.
- Canonical URL −4 — Canonical link is set.
Craft we looked for and did not find
These are the actionable half of the report. None of them are hard, and every one of them would have moved the number.
- Skip-to-content link
- prefers-reduced-motion honoured
- Self-hosted typography
- CSS custom property system
- Hand-authored stylesheet
- No utility-class soup
What this scan cannot see
One request, no JavaScript executed. So everything below is invisible to us, and a page can be excellent or awful at all of it without moving the score:
- Domain registration age — needs WHOIS.
- Whether the nav's /about and /pricing actually resolve — needs a crawl, not one fetch.
- package.json dependency count, orphaned code, clashing abstractions, test coverage — needs the repo.
- Backend security shortcuts: string-built SQL, verify=False, missing auth on endpoints — needs the source.
- Whether row-level security or Firebase rules are actually switched on — needs the project.
- Console errors, hydration warnings and runtime 404s — needs a browser executing the page.
- Real Lighthouse numbers and horizontal scroll at 375px — needs a rendered viewport.
- Whether a JS-handled form posts anywhere — needs a submitted request.