Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Environment config, .env management.
.claude/skills/valec3-backend-deployment-configuration/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-14 | ✗→✓ | ▲ Improved | -10% | 0% |
| case-15 | ✗→✓ | ▲ Improved | -27% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -31% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -37% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -62% | 0% |
bash# .env CI_ENVIRONMENT = production database.default.hostname = localhost database.default.database = myapp database.default.username = dbuser database.default.password = secret database.default.DBDriver = MySQLi JWT_SECRET = your-secret-key API_KEY = external-api-key
php<?php $dbHost = env('database.default.hostname'); $jwtSecret = env('JWT_SECRET');
php<?php $required = ['JWT_SECRET', 'database.default.password']; foreach ($required as $key) { if (!env($key)) { throw new \RuntimeException("Missing env var: {$key}"); } }
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-03 | pass→pass | 10,691 | 6,005 | -44% | 1 | 1 | 0% | 1,834 | 1,524 | -17% | 0 | 0 | — |
case-13 | pass→pass | 9,588 | 2,115 | -78% | 1 | 1 | 0% | 1,732 | 657 | -62% | 0 | 0 | — |
case-14 | fail→pass | 14,619 | 9,968 | -32% | 1 | 1 | 0% | 2,410 | 2,177 | -10% | 0 | 0 | — |
case-15 | fail→pass | 3,284 | 1,104 | -66% | 1 | 1 | 0% | 564 | 414 | -27% | 0 | 0 | — |
case-02 | fail→pass | 10,573 | 6,783 | -36% | 1 | 1 | 0% | 2,215 | 1,525 | -31% | 0 | 0 | — |
case-01 | fail→fail | 14,271 | 9,538 | -33% | 1 | 1 | 0% | 2,791 | 2,241 | -20% | 0 | 0 | — |
case-04 | pass→pass | 10,668 | 8,893 | -17% | 1 | 1 | 0% | 2,209 | 2,028 | -8% | 0 | 0 | — |
case-05 | pass→pass | 6,184 | 4,623 | -25% | 1 | 1 | 0% | 1,215 | 1,162 | -4% | 0 | 0 | — |
case-06 | pass→pass | 10,313 | 5,348 | -48% | 1 | 1 | 0% | 2,150 | 1,371 | -36% | 0 | 0 | — |
case-07 | fail→pass | 4,980 | 1,480 | -70% | 1 | 1 | 0% | 829 | 526 | -37% | 0 | 0 | — |
case-08 | pass→pass | 11,615 | 4,316 | -63% | 1 | 1 | 0% | 2,120 | 982 | -54% | 0 | 0 | — |
case-09 | fail→pass | 9,312 | 2,176 | -77% | 1 | 1 | 0% | 1,856 | 696 | -63% | 0 | 0 | — |
case-10 | fail→pass | 9,581 | 4,724 | -51% | 1 | 1 | 0% | 1,812 | 1,088 | -40% | 0 | 0 | — |
case-11 | fail→pass | 6,566 | 1,803 | -73% | 1 | 1 | 0% | 1,164 | 573 | -51% | 0 | 0 | — |
case-12 | fail→pass | 8,706 | 3,793 | -56% | 1 | 1 | 0% | 1,598 | 976 | -39% | 0 | 0 | — |
case-16 | fail→pass | 7,000 | 2,237 | -68% | 1 | 1 | 0% | 1,357 | 664 | -51% | 0 | 0 | — |
case-17 | fail→pass | 5,596 | 1,607 | -71% | 1 | 1 | 0% | 928 | 568 | -39% | 0 | 0 | — |
case-18 | fail→pass | 8,321 | 1,693 | -80% | 1 | 1 | 0% | 1,072 | 554 | -48% | 0 | 0 | — |
case-19 | fail→fail | 10,081 | 6,223 | -38% | 1 | 1 | 0% | 1,769 | 1,297 | -27% | 0 | 0 | — |
case-20 | fail→pass | 6,174 | 2,950 | -52% | 1 | 1 | 0% | 1,248 | 811 | -35% | 0 | 0 | — |
case-21 | fail→pass | 8,861 | 6,200 | -30% | 1 | 1 | 0% | 1,645 | 1,430 | -13% | 0 | 0 | — |
case-22 | pass→pass | 7,622 | 3,062 | -60% | 1 | 1 | 0% | 1,309 | 800 | -39% | 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 +59 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.