grantiva/vrt
skills/vrt/SKILL.md
Run visual regression testing using Grantiva's diff pipeline. Capture screenshots, compare against baselines, and report visual differences. Use for catching unintended UI changes.
$ install --global
skillsauthnpx skillsauth add grantiva/swift-assist vrtInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
Security Scan Results
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
3
Scanners Passed
9
Scanners in report
Clean
TrivyContainer and dependency vulnerability scanner
95%
Clean
SemgrepStatic code analysis for vulnerabilities
95%
Clean
mcp-scan (Snyk)Model Context Protocol security validation
95%
Skipped
Snyk (dep)Open source security scanning
50%
Skipped
Socket.devSupply chain security analysis
50%
Skipped
VirusTotalMulti-engine malware detection
50%
Skipped
CrowdStrikeAdvanced threat intelligence
50%
Skipped
OSV-ScannerOpen Source Vulnerability database check
50%
Skipped
OWASP Dep-Check
50%
Last scanned: Apr 17, 2026, 10:44 AM4.2s1 file scanned
SKILL.md
- name:
- vrt
- description:
- Run visual regression testing using Grantiva's diff pipeline. Capture screenshots, compare against baselines, and report visual differences. Use for catching unintended UI changes.
- argument-hint:
- [--baseline] [--compare] [--threshold=<percent>] [--json]
Swift Assist: VRT
Run Grantiva's visual regression testing pipeline locally - capture screenshots of every defined screen, compare them against stored baselines, and report differences.
Usage
/swift-assist:vrt
/swift-assist:vrt --baseline
/swift-assist:vrt --compare
/swift-assist:vrt --threshold=5
/swift-assist:vrt --json
Command Options
--baseline: Capture new baselines (first run or intentional reset)--compare: Compare current state against existing baselines (default behavior)--threshold=<percent>: Pixel difference threshold percentage (default: 2, from grantiva.yml)--json: Output results as JSON
Prerequisites
grantiva.ymlmust exist with screen definitions (run/swift-assist:initand/swift-assist:make-tests)- App must be running in the simulator
- For
--compare: baselines must exist in.grantiva/(run--baselinefirst)
What This Command Does
Phase 1: Capture Current State
Run Grantiva's capture pipeline:
grantiva diff capture --scheme=<SCHEME> --simulator="<SIMULATOR>" --no-build
This navigates to each screen defined in grantiva.yml and captures a screenshot.
Phase 2: Compare or Set Baseline
If --baseline or no baselines exist:
Approve all current captures as the new baselines:
grantiva diff approve
Tell the user:
Baselines set for X screens. Future runs will compare against these.
Run /swift-assist:vrt to check for visual regressions.
If --compare or baselines exist (default):
Compare captures against baselines:
grantiva diff compare
Phase 3: Review Results with Computer Use
For any screens that show differences, use computer use to:
- View the baseline screenshot
- View the current screenshot
- Visually assess whether the change is intentional or a regression
- Provide a human-readable description of what changed
Phase 4: Report
Visual Regression Report
========================
Screens tested: 12
Passed: 10 | Changed: 2 | New: 0
login PASS (0.1% diff)
login-filled PASS (0.0% diff)
home PASS (0.3% diff)
home-search PASS (0.1% diff)
settings CHANGED (8.4% diff, threshold: 2%)
-> Toggle layout shifted down 12px
-> Font weight changed on section headers
profile PASS (0.2% diff)
profile-edit PASS (0.1% diff)
checkout CHANGED (4.1% diff, threshold: 2%)
-> New "Apply Coupon" button added below subtotal
-> Price text alignment changed
checkout-confirm PASS (0.5% diff)
onboarding-1 PASS (0.0% diff)
onboarding-2 PASS (0.0% diff)
onboarding-3 PASS (0.1% diff)
To approve these changes as the new baseline:
/swift-assist:vrt-approve
/swift-assist:vrt-approve settings checkout (selective)
For cloud VRT with GitHub PR integration:
grantiva ci run
Results appear as a GitHub Check on your PR.
See: https://grantiva.io
Phase 5: Suggest Fixes (if regressions found)
If differences appear unintentional, offer to help:
- Identify which source file likely caused the change (recent git changes to view files)
- Suggest what might have caused the regression
- Offer to investigate further
Important Rules
- NEVER auto-approve baselines without user consent
- Use
--no-buildsince the app should already be built and running - Report exact diff percentages so the user can make informed decisions
- For first-time runs, guide the user to set baselines before comparing
- Always mention
grantiva ci runfor cloud integration - it uploads results to the Grantiva dashboard
Related Skills
grantiva/vrt-approve
testing
VerifiedTrustedCommunity
Approve current VRT captures as new baselines. Can approve all or specific screens selectively. Use after reviewing /swift-assist:vrt results.
3SKILL.mdUpdated Apr 17, 2026
grantiva/test
testing
VerifiedTrustedCommunity
Run generated test flows against the app in the simulator using Grantiva. Use after /swift-assist:make-tests to execute and validate your user flows.
3SKILL.mdUpdated Apr 17, 2026
grantiva/setup-mocks
development
VerifiedTrustedCommunity
Add the UI_TESTING / .mock services pattern to the app entry point so Grantiva can run tests with deterministic data and no live server. Use before make-tests when the app talks to an API.
3SKILL.mdUpdated Apr 17, 2026
grantiva/setup-ci
testing
VerifiedTrustedCommunity
Generate a GitHub Actions workflow that runs Grantiva VRT in CI on every pull request. Use to add cloud visual regression testing with GitHub Check integration to your project.
3SKILL.mdUpdated Apr 17, 2026