- Home
- Skills
- Testing & Debugging
- ♿ a11y Regression Test Generator
More screenshots
Works with the AI Agents you already use
♿ a11y Regression Test Generator
Generate preview-first Playwright and axe accessibility regression tests for supplied HTML or rendered routes, with source hints and manual-review tasks. Writes only after confirmation; does not certify WCAG compliance.
$15
♿ a11y Regression Test Generator
Example session with this skill installed
Using the supplied synthetic clean and risky HTML fixtures, preview accessibility regression tests. After I approve a fresh output folder, generate the tests and run them in the configured Playwright project. Separate source hints from actual browser results.
- Read your context and instructions
- Compiled the a11y regression test
Representative creator-host validation
This summary describes the synthetic validation run, not a customer audit or a guarantee for every agent host.
| Stage | Observed result |
|---|---|
| Preview | No files written |
| Confirmed generation | Eight output files; reported paths matched disk |
| Clean fixtures | Both passed the generated real-axe tests |
| Deliberately risky fixtures | Both failed as expected |
| Write safety | Existing-directory and unsafe-path refusals passed |
The generated suite exited with a failure because the risky fixtures contained intended defects. That is the expected test outcome, not a clean accessibility verdict.
Source hints and browser findings remain separate. Keyboard use, visible focus, assistive technology, dynamic states and malformed-source edge cases still require review. Next: inspect the findings, make authorized application changes, and rerun the same tests. Agensi supporting-file delivery for this update must be verified separately.
Connects securely to your tools. The creator never sees your data.
About this skill
Generate accessibility regression tests you can run
Turn supplied HTML or explicitly selected rendered routes into Playwright + axe test specs and remediation review notes. Preview first, then create outputs only after confirmation in a fresh directory. Generation is not execution.
Inputs and outputs
Provide individual HTML files, or HTTP(S) routes in an authorized test environment. Component source needs a rendered route and remains unverified context. Outputs are application-specific TypeScript test specs, source-level findings and manual-review tasks. Generated browser tests require successful HTTP navigation before running real axe.
Requirements and delivery
Python 3.10+ runs the standard-library helpers. Running the generated tests also requires Node, @playwright/test, @axe-core/playwright and a matching Playwright browser in your project. Dependencies are not installed automatically. Agensi access uses your connected agent, not a buyer ZIP download. Supporting-file execution depends on the permitted connector workflow; do not export protected skill files as a workaround.
Evidence and limits
Creator-host synthetic testing used real Playwright and axe: two clean fixtures passed and two deliberately risky fixtures failed as expected. This is not proof of compatibility in every host. Source hints do not implement a browser DOM: malformed markup, external CSS, framework syntax and dynamic states need rendered confirmation. Keyboard operation, focus visibility and screen-reader usability require separate review. No WCAG certification or automatic application repair is claimed.
Safe scope
Reads supplied inputs; previews without writing. Confirmed generation refuses overwrites and linked output paths. No deployment, purchases or account changes. Existing Agensi buyer terms apply; the included reference notice grants no additional rights.
Source hints are not a browser DOM or WCAG certification. Malformed HTML, external CSS, framework syntax and dynamic states can be missed or misclassified. Component-to-route associations are unverified. Real axe requires project dependencies and a matching browser; keyboard, focus and assistive technology need manual review. Agensi creator-connected execution passed on synthetic HTML, not an independent buyer account. Live HTTP routes, other browsers and other agent hosts were not tested in that delivery run. Exact-string route dedup can retain equivalent URL variants. Use relative input paths for readable generated filenames.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 21 days ago
Requires
Permissions
File Scopes
Source inspection is read-only. Generated files require confirmation and a fresh output directory. Browser and network access apply only when executing generated tests against explicitly authorized routes; helpers install nothing. No protected skill export, application edits or deployment.