Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Pattern library discovery for pattern-first development. Use BEFORE implementing any new feature, creating components, writing API routes, or adding database operations. Ensures existing patterns are checked first.
.claude/skills/bybren-llc-pattern-discovery/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | -18% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -44% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -31% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -42% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -57% | 0% |
Enforce pattern-first development by checking the pattern library before implementing new functionality.
ALWAYS check patterns before writing new code:
bashls patterns_library/api/ # API route patterns ls patterns_library/ui/ # UI component patterns ls patterns_library/database/ # Database patterns ls patterns_library/testing/ # Testing patterns
See patterns_library/README.md for the complete index.
If pattern exists: Copy, customize, validate.
If missing: Search codebase, implement following conventions, report gap.
| If you need to... | Use this pattern | | ---------------------------------- | --------------------------------- | | Create authenticated API endpoint | api/user-context-api.md | | Create admin-only API endpoint | api/admin-context-api.md | | Handle external webhooks | api/webhook-handler.md | | Validate API input with Zod | api/zod-validation-api.md | | Create protected page | ui/authenticated-page.md | | Build form with validation | ui/form-with-validation.md | | Add new table with RLS | database/rls-migration.md | | Test API endpoints | testing/api-integration-test.md |
patterns_library/README.mdrls-patterns skillfrontend-patterns skill| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 11,691 | 5,639 | -52% | 1 | 1 | 0% | 1,958 | 665 | -66% | 0 | 0 | — |
case-02 | fail→fail | 15,001 | 13,813 | -8% | 1 | 1 | 0% | 2,658 | 2,592 | -2% | 0 | 0 | — |
case-03 | fail→pass | 14,472 | 8,564 | -41% | 1 | 1 | 0% | 2,364 | 1,944 | -18% | 0 | 0 | — |
case-04 | fail→pass | 8,893 | 3,904 | -56% | 1 | 1 | 0% | 1,463 | 816 | -44% | 0 | 0 | — |
case-05 | fail→pass | 7,225 | 2,364 | -67% | 1 | 1 | 0% | 1,086 | 748 | -31% | 0 | 0 | — |
case-06 | fail→pass | 10,857 | 3,187 | -71% | 1 | 1 | 0% | 1,663 | 971 | -42% | 0 | 0 | — |
case-07 | fail→pass | 13,984 | 3,339 | -76% | 1 | 1 | 0% | 2,116 | 915 | -57% | 0 | 0 | — |
case-08 | fail→pass | 17,085 | 2,829 | -83% | 1 | 1 | 0% | 3,299 | 829 | -75% | 0 | 0 | — |
case-09 | fail→pass | 5,367 | 1,553 | -71% | 1 | 1 | 0% | 751 | 659 | -12% | 0 | 0 | — |
case-10 | fail→pass | 6,854 | 3,071 | -55% | 1 | 1 | 0% | 882 | 905 | +3% | 0 | 0 | — |
case-11 | pass→pass | 9,835 | 2,637 | -73% | 1 | 1 | 0% | 1,451 | 818 | -44% | 0 | 0 | — |
case-12 | fail→pass | 12,389 | 2,617 | -79% | 1 | 1 | 0% | 1,820 | 859 | -53% | 0 | 0 | — |
case-13 | pass→pass | 12,506 | 1,775 | -86% | 1 | 1 | 0% | 1,780 | 695 | -61% | 0 | 0 | — |
case-14 | fail→pass | 9,725 | 1,393 | -86% | 1 | 1 | 0% | 1,463 | 625 | -57% | 0 | 0 | — |
case-15 | fail→pass | 10,625 | 1,911 | -82% | 1 | 1 | 0% | 1,610 | 748 | -54% | 0 | 0 | — |
case-16 | fail→pass | 7,985 | 1,855 | -77% | 1 | 1 | 0% | 1,330 | 704 | -47% | 0 | 0 | — |
case-17 | fail→pass | 9,866 | 2,697 | -73% | 1 | 1 | 0% | 1,428 | 825 | -42% | 0 | 0 | — |
case-18 | fail→pass | 6,087 | 2,161 | -64% | 1 | 1 | 0% | 857 | 735 | -14% | 0 | 0 | — |
case-19 | fail→fail | 13,979 | 18,423 | +32% | 1 | 1 | 0% | 2,190 | 614 | -72% | 0 | 0 | — |
case-20 | pass→pass | 3,236 | 3,082 | -5% | 1 | 1 | 0% | 502 | 879 | +75% | 0 | 0 | — |
case-21 | pass→pass | 14,699 | 14,058 | -4% | 1 | 1 | 0% | 2,401 | 2,587 | +8% | 0 | 0 | — |
case-22 | pass→pass | 6,670 | 2,705 | -59% | 1 | 1 | 0% | 969 | 918 | -5% | 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, and 20 counted toward the lift figure. The other 2 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +64 percentage points is the difference between those two pass rates over the 20 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.