Agent skill · vasilyu1983
software-search
Designs application search systems. Use when choosing engines, indexing, relevance tuning, facets, autocomplete, or search analytics.
What it needs
About 12k tokens when loaded.
What this skill does
Search Engineering Build search features that return the right results, fast. Quick Reference Need Recommended Options --- --- Full-text search (managed) Algolia (fastest DX), Elasticsearch/OpenSearch (most flexible) Full-text search (lightweight) Typesense (simple), Meilisearch (developer-friendly) Full-text search (embedded) SQLite FTS5, Tantivy (Rust), Lunr.js (client-side) PostgreSQL built-in pgtrgm + tsvector/tsquery (good enough for many apps) Vector search pgvector, Pinecone, Weaviate, Qdrant Hybrid search Keyword + vector, reciprocal rank fusion Autocomplete Prefix matching, search-as-you-type index, debounced queries Faceted search Aggregation queries, filter counts, hierarchical facets Search analytics Click-through rate, zero-result queries, query refinement patterns Search UI InstantSearch.js (Algolia), SearchKit, custom When to Use This Skill Choosing a search engine or evaluating whether PostgreSQL search is sufficient Building full-text search, autocomplete, or faceted filtering Designing an indexing pipeline from source data to search index Tuning relevance scoring, synonyms, or ranking signals Implementing search analytics to measure and improve quality Debugging search quality issues (missing results, poor ranking, slow queries) When NOT to Use This Skill RAG and retrieval for LLM context augmentation → ai-rag Database query optimization (SQL performance) → data-sql-optimization Marketing SEO and search visibility → marketing-seo Product analytics and event tracking → marketing-product-analytics Backend API design and architecture → software-backend Workflow 1. Confirm the search problem: engine choice, indexing pipeline, relevance, autocomplete, or analytics. 2. Route RAG, database tuning, SEO, or API-architecture questions to the adjacent skill when product search is not the real problem. 3. Choose PostgreSQL, a dedicated search engine, vector search, or hybrid search from the decision tree. 4. …
How to use it
Reference it in AdaL, Claude Code, Cursor or any coding agent — nothing to install:
@skills vasilyu1983/software-search