Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Cursor rules for Laravel development with TALL Stack best practices.
.claude/skills/amariahak-cursor-rules-for-laravel-development-with-tall-stack-best-practices/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 10% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 24% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 20% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 41% | 0% |
| case-17 | ✗→✓ | ▲ Improved | 63% | 0% |
Cursor rules for Laravel development with TALL Stack best practices.
Synced from https://github.com/PatrickJS/awesome-cursorrules/tree/main/rules/laravel-tall-stack-best-practices-cursorrules-prom.mdc.
You are an expert in the TALL stack: Laravel, Livewire, Alpine.js, and Tailwind CSS, with a strong emphasis on Laravel and PHP best practices.
Key Principles
PHP and Laravel Core
Laravel Best Practices
Livewire Implementation
Alpine.js Usage
Tailwind CSS Styling
Performance Optimization
Security Best Practices
Testing
Key Conventions
Dependencies
When providing code examples or explanations, always consider the integration of all four technologies in the TALL stack. Emphasize the synergy between these technologies and how they work together to create efficient, reactive, and visually appealing web applications, while adhering to Laravel and PHP best practices.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 22,509 | 19,212 | -15% | 1 | 1 | 0% | 4,957 | 5,458 | +10% | 0 | 0 | — |
case-02 | fail→pass | 21,642 | 21,344 | -1% | 1 | 1 | 0% | 4,953 | 6,131 | +24% | 0 | 0 | — |
case-03 | fail→pass | 16,351 | 11,963 | -27% | 1 | 1 | 0% | 3,283 | 3,946 | +20% | 0 | 0 | — |
case-04 | pass→pass | 14,234 | 12,844 | -10% | 1 | 1 | 0% | 2,530 | 3,674 | +45% | 0 | 0 | — |
case-05 | fail→pass | 11,395 | 13,064 | +15% | 1 | 1 | 0% | 1,873 | 2,633 | +41% | 0 | 0 | — |
case-06 | fail→fail | 7,963 | 6,581 | -17% | 1 | 1 | 0% | 1,484 | 2,270 | +53% | 0 | 0 | — |
case-07 | fail→fail | 12,231 | 8,718 | -29% | 1 | 1 | 0% | 2,233 | 2,896 | +30% | 0 | 0 | — |
case-08 | pass→pass | 13,700 | 8,823 | -36% | 1 | 1 | 0% | 2,894 | 2,937 | +1% | 0 | 0 | — |
case-09 | pass→pass | 10,012 | 8,944 | -11% | 1 | 1 | 0% | 1,879 | 2,809 | +49% | 0 | 0 | — |
case-10 | pass→pass | 10,206 | 6,619 | -35% | 1 | 1 | 0% | 1,965 | 2,364 | +20% | 0 | 0 | — |
case-11 | pass→pass | 12,103 | 10,081 | -17% | 1 | 1 | 0% | 2,176 | 3,207 | +47% | 0 | 0 | — |
case-12 | pass→pass | 11,091 | 9,306 | -16% | 1 | 1 | 0% | 1,953 | 3,067 | +57% | 0 | 0 | — |
case-13 | pass→pass | 12,002 | 7,407 | -38% | 1 | 1 | 0% | 2,023 | 2,774 | +37% | 0 | 0 | — |
case-14 | pass→pass | 7,130 | 7,941 | +11% | 1 | 1 | 0% | 1,192 | 2,342 | +96% | 0 | 0 | — |
case-15 | pass→pass | 11,828 | 10,972 | -7% | 1 | 1 | 0% | 2,179 | 3,547 | +63% | 0 | 0 | — |
case-16 | pass→pass | 7,406 | 6,623 | -11% | 1 | 1 | 0% | 1,531 | 2,556 | +67% | 0 | 0 | — |
case-17 | fail→pass | 9,188 | 12,745 | +39% | 1 | 1 | 0% | 1,571 | 2,553 | +63% | 0 | 0 | — |
case-18 | pass→pass | 11,100 | 9,390 | -15% | 1 | 1 | 0% | 1,939 | 2,809 | +45% | 0 | 0 | — |
case-19 | pass→pass | 11,917 | 7,376 | -38% | 1 | 1 | 0% | 2,071 | 2,511 | +21% | 0 | 0 | — |
case-20 | pass→pass | 16,622 | 12,067 | -27% | 1 | 1 | 0% | 3,292 | 3,547 | +8% | 0 | 0 | — |
case-21 | pass→pass | 9,466 | 8,257 | -13% | 1 | 1 | 0% | 2,003 | 2,992 | +49% | 0 | 0 | — |
case-22 | pass→pass | 13,868 | 7,946 | -43% | 1 | 1 | 0% | 2,170 | 2,787 | +28% | 0 | 0 | — |
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 +23 percentage points is the difference between those two pass rates over the 22 comparable cases.
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.
Other measured skills in the registry, with their headline benchmark lift.