Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Expert patterns for Algolia search implementation, indexing strategies, React InstantSearch, and relevance tuning Use when: adding search to, algolia, instantsearch, search api, search functionality.
.claude/skills/davila7-algolia-search/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-08 | ✗→✓ | ▲ Improved | -30% | 0% |
| case-19 | ✗→✓ | ▲ Improved | -43% | 0% |
| case-23 | ✗→✓ | ▲ Improved | -15% | 0% |
| case-06 | ✓→✓ | = Same ✓ | -24% | 0% |
| case-02 | ✓→✓ | = Same ✓ | -13% | 0% |
Modern React InstantSearch setup using hooks for type-ahead search.
Uses react-instantsearch-hooks-web package with algoliasearch client. Widgets are components that can be customized with classnames.
Key hooks:
SSR integration for Next.js with react-instantsearch-nextjs package.
Use <InstantSearchNext> instead of <InstantSearch> for SSR. Supports both Pages Router and App Router (experimental).
Key considerations:
Indexing strategies for keeping Algolia in sync with your data.
Three main approaches:
Best practices:
| Issue | Severity | Solution | |-------|----------|----------| | Issue | critical | See docs | | Issue | high | See docs | | Issue | medium | See docs | | Issue | medium | See docs | | Issue | medium | See docs | | Issue | medium | See docs | | Issue | medium | See docs | | Issue | medium | See docs |
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 16,144 | 15,262 | -5% | 1 | 1 | 0% | 3,336 | 3,558 | +7% | 0 | 0 | — |
case-06 | pass→pass | 8,704 | 4,052 | -53% | 1 | 1 | 0% | 1,529 | 1,164 | -24% | 0 | 0 | — |
case-02 | pass→pass | 5,715 | 2,940 | -49% | 1 | 1 | 0% | 1,085 | 940 | -13% | 0 | 0 | — |
case-03 | pass→pass | 6,395 | 1,767 | -72% | 1 | 1 | 0% | 1,270 | 694 | -45% | 0 | 0 | — |
case-04 | pass→pass | 6,309 | 3,024 | -52% | 1 | 1 | 0% | 1,178 | 901 | -24% | 0 | 0 | — |
case-05 | pass→pass | 3,591 | 3,624 | +1% | 1 | 1 | 0% | 635 | 1,069 | +68% | 0 | 0 | — |
case-07 | fail→fail | 7,850 | 5,874 | -25% | 1 | 1 | 0% | 1,515 | 1,513 | -0% | 0 | 0 | — |
case-08 | fail→pass | 7,820 | 3,197 | -59% | 1 | 1 | 0% | 1,493 | 1,050 | -30% | 0 | 0 | — |
case-09 | pass→pass | 4,569 | 2,388 | -48% | 1 | 1 | 0% | 797 | 880 | +10% | 0 | 0 | — |
case-10 | pass→pass | 6,286 | 4,201 | -33% | 1 | 1 | 0% | 1,066 | 1,146 | +8% | 0 | 0 | — |
case-11 | pass→pass | 8,114 | 3,584 | -56% | 1 | 1 | 0% | 1,494 | 1,063 | -29% | 0 | 0 | — |
case-12 | pass→pass | 8,799 | 4,564 | -48% | 1 | 1 | 0% | 1,747 | 1,198 | -31% | 0 | 0 | — |
case-13 | pass→pass | 5,753 | 6,798 | +18% | 1 | 1 | 0% | 1,042 | 1,704 | +64% | 0 | 0 | — |
case-14 | pass→pass | 19,523 | 3,025 | -85% | 1 | 1 | 0% | 1,886 | 909 | -52% | 0 | 0 | — |
case-15 | pass→pass | 10,685 | 1,543 | -86% | 1 | 1 | 0% | 1,827 | 649 | -64% | 0 | 0 | — |
case-16 | pass→pass | 13,479 | 8,952 | -34% | 1 | 1 | 0% | 2,195 | 1,873 | -15% | 0 | 0 | — |
case-17 | pass→pass | 10,913 | 8,532 | -22% | 1 | 1 | 0% | 2,030 | 1,897 | -7% | 0 | 0 | — |
case-18 | pass→pass | 12,130 | 8,186 | -33% | 1 | 1 | 0% | 2,053 | 1,715 | -16% | 0 | 0 | — |
case-19 | fail→pass | 6,803 | 1,555 | -77% | 1 | 1 | 0% | 1,128 | 644 | -43% | 0 | 0 | — |
case-20 | pass→pass | 8,488 | 4,110 | -52% | 1 | 1 | 0% | 1,496 | 1,197 | -20% | 0 | 0 | — |
case-21 | pass→pass | 13,858 | 11,583 | -16% | 1 | 1 | 0% | 2,684 | 2,523 | -6% | 0 | 0 | — |
case-22 | pass→pass | 8,548 | 6,085 | -29% | 1 | 1 | 0% | 1,633 | 1,464 | -10% | 0 | 0 | — |
case-23 | fail→pass | 14,712 | 10,829 | -26% | 1 | 1 | 0% | 2,990 | 2,554 | -15% | 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. 23 cases were attempted. The headline lift of +13 percentage points is the difference between those two pass rates over the 23 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.