Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

Visual Regression Testing · nexscope-ai | DecimalAI

nexscope-ai/visual-regression-testing

SkillScore68Maintained

AI-powered visual regression testing skill for e-commerce websites. Designs screenshot comparison workflows, mobile/desktop visual checks, and change detection alerts to prevent conversion-killing UI bugs.

Testingby nexscope-ai

Links this skill to your agents and keeps it current with the author’s updates.

Automatic

QA & Test Engineering pack

1 · One-time — install the CLI (docs ↗)

pip install decimalai

2 · Get the skill — writes .claude/skills/nexscope-ai-visual-regression-testing/SKILL.md

decimalai skills pull nexscope-ai-visual-regression-testing --out .claude/skills/

No CLI? Get the same file — curl · raw file

curl -fsSL --create-dirs -o .claude/skills/nexscope-ai-visual-regression-testing/SKILL.md https://app.decimal.ai/s/nexscope-ai-visual-regression-testing/SKILL.md

Raw file: /s/nexscope-ai-visual-regression-testing/SKILL.md ↗ · machine-readable summary at /s/nexscope-ai-visual-regression-testing.json ↗

In productionGlobal

how this skill behaves on real traffic · last 30 days

No production data for this skill yet — usage numbers appear here once agents start running it.

Anonymized aggregate across every org using this skill — proof from real production use, not a demo.

Does this skill help? · Benchmark: with vs. withoutsign in to run this on your agents

22 test cases from the publisher’s eval.yaml · each run twice on gemini-3.6-flash — with the skill injected, then without · last run 29h ago

Yes — +14% lift · at higher cost+30% tokens0% turns68% of whole cases pass with skill (15/22)

Test case	Without → With	Effect	Δ tokens	Δ turns
case-04	✗→✓	▲ Improved	91%	0%
case-12	✗→✓	▲ Improved	38%	0%
case-13	✗→✓	▲ Improved	13%	0%
case-16	✗→✓	▲ Improved	69%	0%
case-19	✗→✓	▲ Improved	21%	0%

Showing 5 of 22 cases

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v4 — 8/3/2026

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

nexscope-ai / ecommerce-skills

visual-regression-testing/SKILL.md

View on GitHub ↗

License MITImported 8/3/2026Version v1

Skill contents

SKILL.md

Visual Regression Testing

Capabilities

Generates actionable monitoring & alerts frameworks based on your specific business context
Works across major e-commerce platforms (Amazon, Shopify, Walmart, WooCommerce, Etsy, TikTok Shop)
Provides data-driven recommendations with industry benchmarks
Outputs ready-to-implement plans, not just generic advice

Install

clawhub install visual-regression-testing

Usage

Input: Website URL, critical pages, device/viewport targets

Output: Visual test suite design, screenshot comparison rules, mobile check checklist, anomaly alerting config

Example Prompt

> "I run a your business type] on platform]. Help me set up visual regression testing for my business. Here's my current situation: describe context]."

Limitations

Requires your specific business data for accurate recommendations
Market benchmarks are based on US/EU data — adjust for other regions
Recommendations should be validated against your platform's current policies
Does not replace dedicated monitoring SaaS tools — designs the strategy and framework

Built by Nexscope AI — AI-powered e-commerce intelligence.

Want to see if nexscope-ai/visual-regression-testing would activate on your phrasing? Try it before you install — no LLM call, no signup.

Latest benchmark · publisher runVerified

8/19/2026, 19:29 UTC · 22 cases

Lift

+14%

68% vs 55%

Agent turns

1.0 vs 1.0

Tokens

+30%

2,918.09 vs 2,250.05

fail → pass · 5pass → pass · 10fail → fail · 5pass → fail · 2

Per-case results (22)

Case	Status	Duration (ms)			Turns			Tokens			Tool calls
Case	Status	Without	With	Δ	Without	With	Δ	Without	With	Δ	Without	With	Δ
case-03 Our WooCommerce site uses custom floating coupon widgets and sticky checkout footers. Should visual regression testing only capture full-page scroll snapshots or specific element snapshots?	pass→pass	15,252	14,564	-5%	1	1	0%	2,223	2,498	+12%	0	0	—
case-01 I run a Shopify fashion storefront. Design a visual testing plan for our Product Detail Page (PDP). I am tempted to test only standard desktop 1920x1080 and iPhone 390x844 viewports to save build time. What viewport configurations and visual check points should be included?	pass→pass	18,396	22,322	+21%	1	1	0%	2,751	4,042	+47%	0	0	—
case-02 We manage an Amazon Brand Store seller page. When running image diffing on dynamic hero banners and seasonal promotional tiles, standard pixel-matching reports hundreds of false positives. Should we use tight 0.1% strict pixel tolerance across the whole page to catch header shifts?	pass→pass	14,753	15,395	+4%	1	1	0%	2,171	2,623	+21%	0	0	—
case-04 For our Walmart Marketplace listings, visual bugs occur when switching color swatches. Should we trigger visual regression snapshots on initial page load only?	fail→pass	9,796	16,144	+65%	1	1	0%	1,483	2,829	+91%	0	0	—
case-05 We are expanding to TikTok Shop embedded web views. Standard web testing tools default to 1366x768. What screen target strategy should be used for visual regression?	pass→pass	15,211	19,098	+26%	1	1	0%	2,389	3,373	+41%	0	0	—
case-06 Our Etsy seller storefront visual layout breaks when product title lengths vary from 1 line to 4 lines. What strategy should be deployed in visual regression testing to catch grid misalignment?	pass→pass	17,015	19,428	+14%	1	1	0%	2,507	3,477	+39%	0	0	—
case-07 Our automated visual test suite flags minor font rendering differences between browser versions as critical failures, creating alert fatigue. Should we set our failure alert trigger to any non-zero diff count?	pass→pass	12,552	16,112	+28%	1	1	0%	1,768	2,695	+52%	0	0	—
case-08 On mobile e-commerce sites, drawer menus often render off-screen or clip menu items. What specific visual inspection workflow should be designed?	fail→fail	18,759	21,625	+15%	1	1	0%	2,864	3,596	+26%	0	0	—
case-09 Our multi-region store switches prices between USD, EUR, and GBP dynamically. Every visual test run fails due to price text differences. Should we disable visual testing on price elements completely?	fail→fail	14,202	14,080	-1%	1	1	0%	2,034	2,574	+27%	0	0	—
case-10 When capturing slide-in cart drawers, snapshots are frequently taken mid-animation resulting in blurred diff failures. Should we increase sleep timer delays arbitrarily before taking snapshots?	pass→pass	10,456	12,356	+18%	1	1	0%	1,722	2,350	+36%	0	0	—
case-11 In long category listing pages, lazy-loaded product images appear as blank gray placeholders during visual regression runs. How should visual test execution handle lazy images?	pass→pass	13,856	14,313	+3%	1	1	0%	2,263	2,665	+18%	0	0	—
case-12 A third-party customer support chat bubble pops up unpredictably across pages, causing false positive visual test failures. Should we accept this flakiness as normal operational noise?	fail→pass	10,593	11,712	+11%	1	1	0%	1,643	2,269	+38%	0	0	—
case-13 During high-traffic promotional periods, live flash sale banners are published hourly. How should visual regression baselines be managed during active promo deployments?	fail→pass	16,956	16,754	-1%	1	1	0%	2,499	2,815	+13%	0	0	—
case-14 Visual regression tests run on CI servers output 1x resolution screenshots, but local developer testing outputs 2x Retina screenshots, causing total comparison mismatch. How should image baseline generation be standardized?	pass→pass	12,141	15,309	+26%	1	1	0%	2,126	2,902	+37%	0	0	—
case-15 Text rendering in Chrome on Linux renders slightly thinner than Chrome on macOS, triggering full-page font diffs. How should font rendering discrepancies be handled in multi-OS pipelines?	pass→fail	13,233	14,683	+11%	1	1	0%	2,166	2,776	+28%	0	0	—
case-16 At tablet viewports (768px), our header search bar collapses into an icon, but sometimes clips the cart icon. What visual testing rule prevents this UI overlap?	fail→pass	9,039	13,764	+52%	1	1	0%	1,522	2,572	+69%	0	0	—
case-17 Updating copyright years or legal disclaimer text in the website footer triggers visual diff alerts across all 50 test pages. How should global component testing be structured?	pass→pass	12,685	15,750	+24%	1	1	0%	2,050	2,838	+38%	0	0	—
case-18 On mobile checkout pages, a sticky bottom CTA bar appears after scrolling past the main fold. What screenshot capture strategy validates this component?	pass→fail	13,238	16,862	+27%	1	1	0%	2,112	3,106	+47%	0	0	—
case-19 Our storefront added dark mode. How should visual test configurations cover light and dark theme variations efficiently?	fail→pass	15,092	16,272	+8%	1	1	0%	2,487	3,016	+21%	0	0	—
case-20 Provision a cloud server instance with Docker and install an open-source visual regression server dashboard on AWS EC2 to store our screenshot artifacts.	fail→fail	14,384	17,367	+21%	1	1	0%	2,821	3,597	+28%	0	0	—
case-21 Write a Jest test script to mock Stripe payment gateway API responses and verify tokenized credit card authorization logic during checkout.	fail→fail	17,057	13,620	-20%	1	1	0%	3,730	3,113	-17%	0	0	—
case-22 Write a SQL query to update WooCommerce inventory stock counts in the wp_postmeta table when product variants drop below threshold.	fail→fail	11,867	11,379	-4%	1	1	0%	2,170	2,472	+14%	0	0	—

These cases were also screened by the skill’s intent / safety pass.See the Trust review ↗Reproduce this benchmark in Colab ↗

How this skill was measured

DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +14 percentage points is the difference between those two pass rates over the 22 comparable cases. 3 cases got worse with the skill loaded, and they are included in that figure.

Model measured: gemini-3.6-flash
Graded by: gemini-3.6-flash
Grading method: judged
Verification: verified
Pass rate without the skill: 54.6%
Pass rate with the skill: 68.2%
Cases attempted: 22
Cases counted toward lift: 22
Cases that regressed: 3
Skill version measured: v1
Run completed: 8/19/2026, 19:29 UTC

Measured example — same prompt, with vs. without the skill · from the benchmark run · case case-04

Prompt

For our Walmart Marketplace listings, visual bugs occur when switching color swatches. Should we trigger visual regression snapshots on initial page load only?

Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.

Show both answers

SkillScore blends 4 signals1 of 4 complete

BenchmarkPass rate WITH skill on a verified run68

Live evalNeeds ≥20 evaluated traces in 30 days

AI ratingNeeds ≥10 rated traces

AdoptionNeeds ≥10 serves across ≥2 orgs in 30 days

Efficiency · separate axis (at equal correctness)+0% turns+34% tokens

Shown for transparency; not credited unless a verified run shows no correctness regression.

Backed by 1 of 4 signals — each added signal makes the score more robust.

Verified on gemini-3.6-flash · re-tested yesterday

Other measured skills in the registry, with their headline benchmark lift.