Reader-First Design Audit

What the company appears to be: A self-reported, single-person project named Reader-First Design Audit, which describes itself as a reusable Codex skill for auditing websites or frontend repositories with an emphasis on readability and structural clarity.

What changed: The author states they iteratively developed this tool from recurring problems in their own products and other interfaces. They built it using Codex and GPT-5.6, incorporating feedback from forward tests against AI-assisted sites and repositories.

Single most important open question: Is there any evidence of actual usage or adoption by others beyond the author's own testing?

Note: This analysis is based entirely on the self-reported description provided by the author. No external verification, traction data, revenue figures, customer names, or third-party sources are available. All claims are treated as stated by the author and not independently confirmed.

The description states that Reader-First Design Audit is a reusable, read-only Codex skill designed to audit public websites or frontend repositories.

It examines:

• Orientation, hierarchy, scanning, and comprehension;
• Typography and reading comfort;
• Muted, thin, undersized, or role-inappropriate text;
• Semantic HTML for people, browsers, search engines, and AI tools;
• Tokens, CSS drift, containers, responsive behavior, and interaction state;
• Directly supported accessibility safeguards without pretending to replace specialist assistive-technology testing.

The default report includes:

• A blunt verdict on whether people can read the interface;
• Three actions by reader impact;
• What to leave alone (strengths and intentional exceptions);
• Whether people and machines understand the structure;
• Receipts — measurements, source ownership, confidence, and limitations.

It is bounded to one deeply inspected route plus up to two propagation checks unless more serious harm requires expansion.

Inference: The product appears to be a tool for automated auditing of web interfaces with focus on readability and structural clarity. It uses AI (Codex + GPT-5.6) to analyze content and generate reports.

The author positions the tool as:

• A blunter tool than typical audits that respond with long checklists or isolated CSS complaints.
• Focused on reader impact, not just technical compliance.
• Designed to answer: Can people read and understand this? What are the three changes that matter most?
• Not a replacement for specialist testing but a focused, high-leverage layer.

Key claims include:

• Produces an actionable 80/20 report instead of exhaustive defect inventories.
• Preserves intentional design character and explicitly says what not to change.
• Supports site-only, source-only, and source-plus-site evidence.
• Separates reader friction from confirmed standards failures, system causes, and recommendations.

Inference: The positioning evolved from a general-purpose audit into a focused readability-focused skill that prioritizes user experience over technical compliance alone. It emphasizes clarity and actionability over completeness.

The description does not clearly define target customers or ideal customer profiles (ICP). However, it implies usage by:

• Developers or designers working on web interfaces.
• Teams building AI-assisted websites or frontend repositories.
• Anyone looking to improve readability and structure in digital products.

It is described as a Codex skill, suggesting integration into AI workflows, particularly those involving developers using tools like GitHub or OpenAI agents.

Inference: Likely targets developers or design teams who work with web interfaces and want to ensure readability and structural clarity. The tool may appeal to those building AI-assisted sites where accessibility is important but not fully covered by existing tools.

There is no evidence of a business model, pricing structure, or monetization strategy in the description.

Not evidenced

The project was built using:

• Agent-skills
• Codex
• GPT-5.6
• OpenAI
• GitHub
• APCA (Advanced Perceptual Contrast Algorithm)
• WCAG 2.2

It is described as a reusable repository with installation instructions.

Key technical details include:

• The skill supports site-only, source-only, and source-plus-site evidence.
• It records both WCAG and APCA separately alongside font, size, weight, role, and reading context.
• Forward-tested against AI-assisted sites, repositories, and a non-AI control.
• Passes Codex’s skill package validator.

Inference: The tool is built using modern AI/ML frameworks (Codex, GPT) and integrates with developer ecosystems like GitHub. It uses structured data to evaluate readability and structure.

The description states:

• The author iteratively developed the skill.
• Forward-tested early versions against real sites and repositories.
• Includes installation, testing, and reviewer guidance in a small reusable repository.
• Submitted to the OpenAI 2026 hackathon on Devpost.

There is no evidence of:

• Revenue
• Customers
• Adoption metrics
• Usage beyond author’s own testing

Not evidenced

The description does not mention competitors or competitive landscape.

Not evidenced

Key risks and red flags based on the self-reported information:

1. No evidence of traction or adoption: The tool is described only as a personal project with no external usage.
2. Single-person development: Only one team member (MV Braverman) is mentioned, raising questions about scalability or long-term maintenance.
3. Limited scope: It focuses on readability and structure but excludes full accessibility testing or assistive technology validation.
4. Self-reported maturity: No independent validation of effectiveness or performance beyond forward tests.
5. Unclear commercial viability: No pricing, monetization, or business model described.

Inference: The project is experimental and self-contained, with no clear path to market traction or commercial viability without further evidence.

1. Has the tool been tested in real-world environments beyond your own?
2. Are there any users or adopters outside of personal testing?
3. What are the plans for scaling beyond a single-person development model?
4. How does this tool differ from existing accessibility or design system audit tools?
5. Is there any plan to monetize or commercialize the skill?
6. What is the long-term vision for expanding its functionality or reach?