similarity-search-patterns — Claude Code skill
This skill helps build similarity search with vector databases for semantic search, RAG retrieval, nearest-neighbor queries, and hybrid search.
/plugin marketplace add wshobson/agents
/plugin install llm-application-dev@claude-code-workflowsWhat similarity-search-patterns does
similarity-search-patterns covers patterns for similarity search in vector databases. It fits semantic search, RAG retrieval, recommendation engines, and nearest-neighbor queries. The material explains cosine, Euclidean, dot-product, and Manhattan distance, plus the trade-offs between Flat, HNSW, and IVF+PQ indexes.
The skill shows how to tune ef_search and nprobe, combine semantic and keyword search, pre-filter the search space, and measure recall. Templates and worked examples are stored in references/details.md. It also covers search latency, P99, and the cost of vector storage.
Install it when a project needs vector retrieval or RAG and the team will measure result quality. A regular full-text search without vector data does not need this skill.
What it helps with
- build semantic search over vector data
- implement nearest-neighbor queries
- choose an index for the dataset size
- combine semantic and keyword results
- tune recall and search latency
Example prompts for the agent
- Design semantic search for this project and choose an appropriate index.
- Compare Flat, HNSW, and IVF+PQ for our use case and explain the trade-offs.
- Add hybrid search with pre-filtering and define how to measure recall.
How to install similarity-search-patterns
1Plugin marketplace
/plugin marketplace add wshobson/agents/plugin install llm-application-dev@claude-code-workflowsThis installs the whole llm-application-dev plugin with all its skills
2Skills CLI (Vercel)
npx skills add wshobson/agents --skill similarity-search-patterns -a claude-code3Manually: put the files in place
~/.claude/skills/similarity-search-patterns/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/similarity-search-patterns/
1Skills CLI (Vercel)
npx skills add wshobson/agents --skill similarity-search-patterns -a codex2Manually: put the files in place
.agents/skills/similarity-search-patterns/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add wshobson/agents --skill similarity-search-patterns -a cursor2Manually: put the files in place
.cursor/skills/similarity-search-patterns/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install wshobson/agents similarity-search-patterns2Manually: put the files in place
.github/skills/similarity-search-patterns/SKILL.mdPlatform CLI
gemini skills install https://github.com/wshobson/agents --path plugins/llm-application-dev/skills/similarity-search-patternsManually: put the files in place
.devin/skills/similarity-search-patterns/SKILL.mdThe legacy .windsurf/skills/ path still works
Claude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
From the same set213
The whole set →Similar
Questions about similarity-search-patterns
How do I install similarity-search-patterns in Claude Code?
Run npx skills add wshobson/agents --skill similarity-search-patterns -a claude-code. The skill file is placed at ~/.claude/skills/similarity-search-patterns/SKILL.md.
What does similarity-search-patterns do?
It provides patterns for similarity search with vector databases. The skill covers distance metrics, index selection, search tuning, hybrid retrieval, and recall checks.
Which platforms support similarity-search-patterns?
The skill is available for Claude Code, Codex, and Cursor. Each platform has its own install command and skill-file path.
Source text
SKILL.md · MIT
--- name: similarity-search-patterns description: Implement efficient similarity search with vector databases. Use when building semantic search, implementing nearest neighbor queries, or optimizing retrieval performance. --- # Similarity Search Patterns Patterns for implementing efficient similarity search in production systems. ## When to Use This Skill - Building semantic search systems - Implementing RAG retrieval - Creating recommendation engines - Optimizing search latency - Scaling to millions of vectors - Combining semantic and keyword search ## Core Concepts ### 1. Distance Metrics | Metric | Formula | Best For | | ------------------ | ------------------ | --------------------- | --- | -------------- | | **Cosine** | 1 - (A·B)/(‖A‖‖B‖) | Normalized embeddings | | **Euclidean (L2)** | √Σ(a-b)² | Raw embeddings | | **Dot Product** | A·B | Magnitude matters | | **Manhattan (L1)** | Σ | a-b | | Sparse vectors | ### 2. Index Types ``` ┌─────────────────────────────────────────────────┐ │ Index Types │ ├─────────────┬───────────────┬───────────────────┤ │ Flat │ HNSW │ IVF+PQ │ │ (Exact) │ (Graph-based) │ (Quantized) │ ├─────────────┼───────────────┼───────────────────┤ │ O(n) search │ O(log n) │ O(√n) │ │ 100% recall │ ~95-99% │ ~90-95% │ │ Small data │ Medium-Large │ Very Large │ └─────────────┴───────────────┴───────────────────┘ ``` ## Templates and detailed worked examples Full template library and detailed worked examples live in `references/details.md`. Read that file when you need the concrete templates. ## Best Practices ### Do's - **Use appropriate index** - HNSW for most cases - **Tune parameters** - ef_search, nprobe for recall/speed - **Implement hybrid search** - Combine with keyword search - **Monitor recall** - Measure search quality - **Pre-filter when possible** - Reduce search space ### Don'ts - **Don't skip evaluation** - Measure before optimizing - **Don't over-index** - Start with flat, scale up - **Don't ignore latency** - P99 matters for UX - **Don't forget costs** - Vector storage adds up
© Seth Hobson · MIT · wshobson/agents
