01 Architect · L0 Mindset
L0 Mindset — opensource framework for accessible thinking
Compliance starts before code. L0 Mindset is the Architect-tier foundation of ariada.ai — an opensource bible (10 design rules, Cobbler's Shoes Test, Progressive Enhancement architecture) plus a brand-bibled LMS for client team enablement. Shipped on GitHub under MIT / Apache. The goal: clients build right from day 1, so the fix backlog never accumulates in the first place.
What it does
-
10 design rules
The canonical 10 commandments of accessible design — semantic HTML first, triple encoding (color + shape + text), Progressive Enhancement, focus discipline, 44×44 touch targets, and more. One short list, internalized by every team member.
-
Cobbler's Shoes Test
One question before every release: can a blind user, a color-blind user, and a keyboard-only user use this feature? If the answer is no, it does not ship. A cultural fence, not a checklist item.
-
Progressive Enhancement architecture
Three-layer rendering model — HTML carries all meaning, CSS enhances presentation, JS adds interactivity. Disabling CSS or JS must not break content or core tasks. Engineering becomes accessible by construction.
-
Brand-bibled LMS
Self-paced training tracks for designers, engineers, and product managers. White-labelled to the customer's brand, grounded in WCAG 2.2 AA + EN 301 549 + EAA framing. Cohort-based onboarding plus reference micro-courses.
-
Behavioral nudges
Choice architecture, identity shift ("I am an Inclusive Designer"), and scaffolded practice groups. Drawn from Fogg behavior model + nudge theory — the goal is to reshape default behaviors, not stack more reminders.
-
Inclusive Designer cert
Identity badge for client team members who complete the curriculum and ship audited work. Mentorship, practice groups, and a community of practice maintain the standard after enrollment ends.
Layer mapping
| Axis | Role | Direction |
|---|---|---|
| L0 Mindset | This layer — opensource framework, training, identity system | Source |
| L1 Design | Designers internalize L0 rules, then enforce them via Figma / Adobe / Sketch / Canva plugins | Feeds |
| L2 Dev Tools | Engineers internalize L0 rules, then enforce them via ESLint / Stylelint / IDE / CI | Feeds |
| L3 Testing (WCAG / EAA) | Scanner-axis runs against shipped product; fewer findings when L0 is internalized | Feeds |
| L5–L10 axes | Security, SEO, perf, sustainability, AI Act, legal — all benefit from upstream-by-construction discipline | Feeds |
Patent posture
L0 Mindset has no patent and no provisional. Psychology, training curricula, behavioral nudges, identity shift programs, and community practice are deliberately a moat-free zone — there are no a11y patents in this space, and there should not be. ARIADA ships L0 as opensource (MIT / Apache) on GitHub and treats it as the gift that compounds client adoption. Monetization happens downstream — via L1 Design plugins, L2 Dev Tools, and the L3–L10 scanner-axes. The bible is free; the loom is the product.
Differentiator
Most accessibility vendors lead with detection ("we'll find the bugs") or remediation ("we'll patch them"). ARIADA leads with prevention — transform the team that builds the product, and detection finds less, and remediation costs less. L0 is where that transformation lives. No competitor ships an opensource framework + brand-bibled LMS at this layer; the closest analogues (Deque University, WAS certification) are paid silos detached from any tooling downstream. ARIADA's L0 is free, opinionated, and integrated with the L1–L10 stack the same team uses to ship.
Cross-references
-
L1 Design — design-time enforcement
Designers who have internalized L0 enforce the same rules inside Figma, Adobe, Sketch, and Canva via the L1 plugin family. See L1 Design →
-
L2 Dev Tools — build-time enforcement
Engineers who have internalized L0 enforce the same rules inside the IDE and CI via ESLint plugins, the
@ariada-org/mcp-server, and the GitHub Action. See L2 Dev Tools → -
Pipeline modules — Detect tier
Scanner-axis modules (J, D, G, C, H, K) measure the shipped product. Fewer findings when L0 is internalized upstream. See all modules →
-
LLM remediation cascade
When findings do appear, the LLM remediation cascade drafts source-level pull requests for the engineering team to review and merge. See LLM remediation cascade →
-
Why ariada.ai
Prevention beats remediation. The triad framing (Architect · Detect · Remediate) starts here. See the why →
-
Layer architecture detail
Full L0–L10 walk-through with patent-status colour-coding for each axis. See Layer architecture →
Source-level remediation only — agentic suggestions are not autonomous deployments; pull requests require client merge. Not a legal certification body. For accredited certification, customers consult a notified body or registered auditor.