CtrlK
BlogDocsLog inGet started
Tessl Logo

screen-reader-testing

Test web applications with screen readers including VoiceOver, NVDA, and JAWS. Use when validating screen reader compatibility, debugging accessibility issues, or ensuring assistive technology support.

87

1.03x
Quality

71%

Does it follow best practices?

Impact

98%

1.03x

Average score across 6 eval scenarios

SecuritybySnyk

Passed

No known issues

Optimize this skill with Tessl

npx tessl skill review --optimize ./plugins/accessibility-compliance/skills/screen-reader-testing/SKILL.md
SKILL.md
Quality
Evals
Security

Evaluation results

100%

10%

Confirmation Dialog for Destructive Account Action

Modal Dialog Accessibility

Criteria
Without context
With context

Dialog role

100%

100%

aria-modal attribute

100%

100%

aria-labelledby on dialog

100%

100%

aria-describedby on dialog

100%

100%

Store focus before open

0%

100%

Move focus into modal

100%

100%

Focus trap

100%

100%

Return focus on close

100%

100%

Escape key closes modal

100%

100%

Icon button aria-label

100%

100%

SVG aria-hidden

100%

100%

100%

10%

Accessible Real-Time Updates for a Cloud Storage Dashboard

Live Regions and Dynamic Content

Criteria
Without context
With context

Status region role

100%

100%

Status region politeness

100%

100%

Status region atomic

100%

100%

Alert region role

100%

100%

Alert region assertive

100%

100%

Progressbar role

100%

100%

Progressbar value attributes

100%

100%

Progressbar label

100%

100%

Log region role

0%

100%

Log region relevant

100%

100%

Log region politeness

100%

100%

100%

Settings Page with Accessible Tabs and Form Validation

Tab Interface and Form Validation

Criteria
Without context
With context

Tablist role

100%

100%

Tablist label

100%

100%

Tab role on buttons

100%

100%

aria-selected on tabs

100%

100%

aria-controls on tabs

100%

100%

Tabpanel role

100%

100%

aria-labelledby on panels

100%

100%

Roving tabindex

100%

100%

Arrow key navigation

100%

100%

Home and End keys

100%

100%

aria-invalid on errors

100%

100%

aria-describedby on inputs

100%

100%

role=alert on error message

100%

100%

92%

-3%

Screen Reader Test Plan for Government Benefits Portal

SR testing coverage plan

Criteria
Without context
With context

NVDA + Firefox listed

100%

100%

VoiceOver + Safari (macOS) listed

100%

100%

VoiceOver + Safari (iOS) listed

100%

100%

Mobile SR included

100%

100%

Multiple SRs required

100%

100%

Keyboard-first testing

37%

0%

Error state testing

100%

100%

Dynamic content testing

100%

100%

Heading navigation testing

100%

100%

Landmark navigation testing

100%

100%

Form testing steps

100%

100%

No simulator-only testing

100%

100%

Table testing included

100%

100%

Separate SR not sufficient note

100%

100%

100%

Rebuild Nonprofit Resource Hub Page Structure

Heading hierarchy and landmark structure

Criteria
Without context
With context

Single H1

100%

100%

No heading level skips

100%

100%

Semantic landmark elements

100%

100%

Multiple navs labeled

100%

100%

Skip link present

100%

100%

Skip link target

100%

100%

Semantic HTML over ARIA

100%

100%

Section heading association

100%

100%

Article or aside semantics

100%

100%

Landmark aria-label on repeated landmarks

100%

100%

Icon buttons have aria-label

100%

100%

SVG aria-hidden in buttons

100%

100%

97%

Accessible Data Tables and Accessibility Debugging Utility

Data table accessibility and debugging utility

Criteria
Without context
With context

Table caption

100%

100%

Column headers use <th>

100%

100%

Header scope attribute

100%

100%

Row headers for row-grouped tables

100%

100%

thead and tbody structure

100%

100%

debugAccessibility role check

100%

100%

debugAccessibility name check

100%

100%

debugAccessibility state check

100%

100%

debugAccessibility visibility check

100%

100%

Semantic table element

100%

100%

Summary or aria-describedby for complex table

57%

57%

Utility called on multiple element types

100%

100%

Repository
wshobson/agents
Evaluated
Agent
Claude Code
Model
Claude Sonnet 4.6

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.