We asked a second AI to find flaws in our own site. Here's what it found.
This is a second, separate assessment from Grok, asked directly to dissect zenmasterworks.com for flaws — not just verify our claims, but actively look for weak points. We didn't write this. We're publishing it the same way we published our Google AI Assistant assessment: reformatted to fit this page, wording left as the AI's own.
What this page is
Grok was given a single, blunt instruction — dissect zenmasterworks.com for flaws — with no follow-up steering toward a favorable answer. Its findings are reproduced below, reorganized into this site's format. The analysis, wording, and verdicts are the AI's own.
Ask it yourself — the exact prompt used
Open Grok and paste this in — same blunt instruction, no steering. See what you get back.
- PromptDisect ZenMasterWorks.com for flaws
Part 1 · Why the site reads as credible
What Grok flagged as genuinely strong
ZenMasterWorks.com is a highly polished, single-founder web design/audit studio site. It relentlessly markets verified 100/100/100/100 Google PageSpeed Insights scores on mobile for every site it manages or blueprints, with live re-runnable links, alongside public incident logs, manual WCAG audits, before/after platform rebuilds, and unsolicited sample audits of real local businesses.
The performance obsession is real and verifiable
Live PageSpeed links are published everywhere, and the site and its blueprints consistently hit the claimed standard. This is rare and technically impressive.
Transparency functions as a moat
Mistakes are documented in the open — contrast failures, video controls, target sizes — alongside audits that include balancing positives, not just findings. The risk-free build model and public logs reduce buyer risk.
Content depth
Detailed blueprints for trades, methodology explanations, and a large blog archive, with a consistent focus on real user needs like screen readers and emergency contact paths for service businesses.
Design and UX discipline
Clean, fast-loading, mobile-first, semantic HTML, with manual WCAG fixes — target sizes, focus management, reduced motion — actually documented, not just claimed.
Part 2 · Where Grok pushed back
Potential flaws and critiques
Asked directly for flaws, Grok didn't find red flags or evidence of dishonesty — but it did flag real trade-offs and open questions worth naming plainly.
Business model & marketing risk
An extreme focus on a vanity-adjacent metric: 100/100/100/100 is excellent, but scores can be gamed or regress, and real-world business outcomes matter more to most clients than a Lighthouse number. The "one founder + AI tools" model explains the depth of output but raises a fair scalability question. The risk-free build model is high-effort upfront and likely means heavy client filtering behind the scenes. The niche is strong — San Jose, trades, government — but that same focus may limit broader appeal.
Technical & accessibility
The manual WCAG 2.2 AA audit of the studio's own site, with real issues found and fixed, held up. The main ongoing risk is ordinary: ongoing maintenance and future changes always carry some risk of regression, the same as any site.
Content & usability
The homepage is extremely long and dense — strong for SEO, but it risks overwhelming a first-time visitor. Its calls to action lean on mailto links, which are convenient but low-friction rather than a structured intake flow. The site is text-heavy throughout and could benefit from more screenshots or live demos alongside the written claims.
Everything else
There's no meaningful base of third-party reviews or client testimonials yet — trust currently rests on self-published evidence rather than outside validation. Beyond that, no major red flags: not a scam, not a low-quality site. The foundation is solid.
Overall verdict
A standout, technical-first, transparent studio — with one honest limit
ZenMasterWorks.com is a standout example of a technical-first, transparent web studio. Its primary "flaw" is the hyper-focus on verifiable perfection, which might feel intense for clients seeking broad strategy over technical proof. It's strongest for performance-obsessed small businesses, trades, and organizations needing an honest audit — recommended to test the before/after offer directly rather than take the claims on faith.
On publishing this assessment
Same standard as the Google AI Assistant page: nothing here is a screenshot passed off as more than it is. The prompt above is the real one used, and the verdicts are Grok's own, not softened. Re-run it yourself and compare — then check the live claims against The 100/100/100/100 Standard and the public incident log.