Basano

See it in action

A verdict, run by the engine

This is a real run of two engines. Basano scopes the review, selects the verification skills each area needs, and synthesizes the verdict below; then Tholo reads the fails and prepares the fix. The findings are a sample, standing in for the live checks that come next, but the engines that read them are real.

The request

What was built
a finished landing page for a new product, with hero, features, testimonials, and a contact form
The critique asked for
UX and UIAccessibilityCore Web VitalsContent and editorial standardsSEO, AEO, and GEO

These are the dimensions asked for. The rows below are these same dimensions, each with what Basano could actually determine. The chips are the ask; the marks are what the engine found.

The standard
production-ready

What Basano did

It scoped the review into the asked dimensions, selected the matching audit skills from the catalog, checked each one, and returned a verdict of what holds and what fails.

What was checked

1 holds2 fails2 not assessable
  • UX and UINot assessable

    This static sample has no rendered capture, so the above-fold layout could not be assessed. The rendered-capture path in a later step is what fills this in.

  • AccessibilityhighFails
  • Core Web VitalsNot assessable

    This sample carries no deployed URL and no performance check has run, so Core Web Vitals could not be measured. The live run in a later step is what fills this in.

  • Content and editorial standardslowFails
  • SEO, AEO, and GEOHolds

    Only SEO was assessed here. AEO and GEO are not yet checked by the runner.

The full verdict

VerdictFails

At least one area fails. A pass is structurally impossible while any fail stands.

Holds
2
Fails
2
Not assessable
2

What fails

  • Color contrast on the hero call to action

    high

    CTA text sits at a 2.8:1 contrast ratio against the hero gradient; WCAG AA requires 4.5:1. Visitors with low vision cannot read the primary action.

  • Feature copy leans on unexplained jargon

    low

    Three of five feature blocks use product-team shorthand with no translation. A first-time visitor will not parse it.

Could not assess

  • Above-fold layout and primary-task prominence

    This static sample has no rendered capture, so the above-fold layout could not be assessed. The rendered-capture path in a later step is what fills this in.

  • Page speed and Core Web Vitals

    This sample carries no deployed URL and no performance check has run, so Core Web Vitals could not be measured. The live run in a later step is what fills this in.

What holds

  • Titles, meta descriptions, headers, and internal links

    On-page SEO held against the standard: a single descriptive title and meta, one H1 with a sensible heading order, and internal links with meaningful anchors.

  • Crawl, indexing, sitemap, and robots

    Technical SEO held against the standard: the page is crawlable and indexable, present in the sitemap, and not blocked by robots.

Where Basano stopped

The critic does not guess. These are the parts it would not turn into a verdict, kept visible rather than folded into a pass.

  • Above-fold layout and primary-task prominence: This static sample has no rendered capture, so the above-fold layout could not be assessed. The rendered-capture path in a later step is what fills this in.
  • Page speed and Core Web Vitals: This sample carries no deployed URL and no performance check has run, so Core Web Vitals could not be measured. The live run in a later step is what fills this in.

Areas and what was checked

  • review above-fold layout and primary-task prominence

    Verification skills selected: competitor-experience-audit

  • review accessibility against WCAG

    Verification skills selected: accessibility-audit

  • review performance and page speed

    Verification skills selected: performance-optimization, seo-onpage, landing-page-copy

  • review content quality and editorial standards

    Verification skills selected: design-standards, content-strategy, editorial-qa, ai-content-collaboration

  • review on-page SEO: titles, meta descriptions, headers, internal links

    Verification skills selected: seo-onpage

  • review technical SEO: crawl, indexing, sitemap, robots

    Verification skills selected: seo-technical

Synthesized summary

Verdict: fails. 2 failing finding(s); 2 holds; 2 not assessable.
Fails (highest severity first):
  - [HIGH] Color contrast on the hero call to action: CTA text sits at a 2.8:1 contrast ratio against the hero gradient; WCAG AA requires 4.5:1. Visitors with low vision cannot read the primary action.
  - [LOW] Feature copy leans on unexplained jargon: Three of five feature blocks use product-team shorthand with no translation. A first-time visitor will not parse it.
Could not assess:
  - Above-fold layout and primary-task prominence: This static sample has no rendered capture, so the above-fold layout could not be assessed. The rendered-capture path in a later step is what fills this in.
  - Page speed and Core Web Vitals: This sample carries no deployed URL and no performance check has run, so Core Web Vitals could not be measured. The live run in a later step is what fills this in.
Holds:
  - Titles, meta descriptions, headers, and internal links: On-page SEO held against the standard: a single descriptive title and meta, one H1 with a sensible heading order, and internal links with meaningful anchors.
  - Crawl, indexing, sitemap, and robots: Technical SEO held against the standard: the page is crawlable and indexable, present in the sitemap, and not blocked by robots.

The findings here are a sample input. The next step runs the assessable checks against a live URL and feeds the engine real observations; the verdict synthesis stays exactly as shown.

Audit your own page

The system

Krine decides, Tholo builds, Basano proves

Three engines in one loop. Select an engine to see its role. Basano and Tholo both ran here, prove then prepare; Krine is the part of the loop this page does not exercise, and nothing executed.

Cogent runtimethe loop closes to one approve step

Basano proves. Proves the build against the standard and reports what holds and what fails.

The next step

Basano judged. Tholo prepared the fix.

Basano stops at the verdict by design; it does not change the build. So the fails flowed to Tholo, the builder, which read them and prepared the work to close each one. That prepared fix is right below.

Tholo prepares; it does not apply. It stops at the execute seam, the line the system does not cross on its own. You, or an agent you authorize, own the apply. The orchestrator does the planning, so what is left is one approve step.

Basano provesTholo prepares the fixyou approve

Prepared fix dispatch

Needs clarification

Tholo's plan for Basano's fails

Generated by the Tholo engine from the two fails above, one step per fail. This is prepared, not run.

  1. 1

    fix the accessibility against WCAG failure, color contrast on the hero call to action

    Skills selected

    accessibility-auditdesign-standards

    8 required inputs, 16 notes in the dispatch below

  2. 2

    fix the content quality and editorial standards failure, feature copy leans on unexplained jargon

    Needs clarification. task matched 6 skills above the relative floor; returning top 5. Narrow the goal for a sharper selection. Tholo will not prepare a dispatch it is not confident in; narrow this goal and re-run.

Execute seam

Prepared, not run. plan prepared; execution is a human action. Run the dispatch yourself once you have approved it. You, or an agent you authorize, apply it once you have approved it. The loop closes to one step the human owns.

The dispatch, ready to hand to your agent

# Prepared fix dispatch

Source: Basano verdict (sample) handed to Tholo
Task: fix the accessibility against WCAG failure, color contrast on the hero call to action and fix the content quality and editorial standards failure, feature copy leans on unexplained jargon
Status: needs clarification

## Step 1: fix the accessibility against WCAG failure, color contrast on the hero call to action
Skills: accessibility-audit (skills/accessibility-audit/SKILL.md), design-standards (skills/design-standards/SKILL.md)
Required inputs:
- The site or product under audit
- The scope (full site, specific section, specific user flow)
- The target standard (WCAG 2.1 AA is most common)
- Any specific concerns or known issues
- Tools available (automated scanners, screen readers, manual testing)
- The page or component being built or reviewed
- The brand's design tokens (colors, type, spacing) - or willingness to define them
- The target devices and viewports
Notes:
- conflict: seo-backlink-audit when-NOT-to-use names seo-keyword-gap-audit; kept higher-scoring seo-backlink-audit
- conflict: seo-backlink-audit when-NOT-to-use names seo-site-health-audit; kept higher-scoring seo-backlink-audit
- conflict: after-action-report when-NOT-to-use names incident-response; kept higher-scoring after-action-report
- conflict: brand-identity when-NOT-to-use names brand-style-guide; kept higher-scoring brand-identity
- conflict: monitoring-and-alerting when-NOT-to-use names after-action-report; kept higher-scoring after-action-report
- conflict: code-review-web when-NOT-to-use names qa-testing; kept higher-scoring code-review-web
- conflict: code-review-web when-NOT-to-use names performance-optimization; kept higher-scoring code-review-web
- conflict: frontend-component-build when-NOT-to-use names code-review-web; kept higher-scoring frontend-component-build
- conflict: vertical-site-conventions when-NOT-to-use names creative-direction; kept higher-scoring vertical-site-conventions
- conflict: vertical-site-conventions when-NOT-to-use names brand-archetype-system; kept higher-scoring vertical-site-conventions
- conflict: vertical-site-conventions when-NOT-to-use names frontend-component-build; kept higher-scoring frontend-component-build
- conflict: creative-brief-selector when-NOT-to-use names logo-design; kept higher-scoring creative-brief-selector
- conflict: creative-brief-selector when-NOT-to-use names brand-identity; kept higher-scoring brand-identity
- conflict: seo-competitor when-NOT-to-use names seo-onpage; kept higher-scoring seo-competitor
- conflict: seo-competitor when-NOT-to-use names seo-content-audit; kept higher-scoring seo-competitor
- boundary: competitor-experience-audit: boundary fired on bullet "Auditing a single page's on-page SEO or accessibility against the audit suite (use `seo-onpage`, `accessibility-audit`)"; score 18.75 below override floor 37.51

## Step 2: fix the content quality and editorial standards failure, feature copy leans on unexplained jargon
Needs clarification: task matched 6 skills above the relative floor; returning top 5. Narrow the goal for a sharper selection.
Tholo will not prepare a dispatch it is not confident in. Narrow this goal and re-run.

## Required inputs (all ready steps)
- withheld until the steps that need clarification are narrowed

## Execute seam
Prepared, not run. plan prepared; execution is a human action. Run the dispatch yourself once you have approved it.
You, or an agent you authorize, apply this once you have approved it.