similarity-search-patterns

Implement efficient similarity search with vector databases. Use when building semantic search, implementing nearest neighbor queries, or optimizing retrieval performance.

Install with Tessl CLI

npx tessl i github:wshobson/agents --skill similarity-search-patterns

What are skills?

Review — 70%

Does it follow best practices?

If you maintain this skill, you can automatically optimize it using the tessl CLI to improve its score:

npx tessl skill review --optimize ./path/to/skill

Learn more

Validation — 10 / 11 Passed

Validation for skill structure

SKILL.md

Review

Evals

Evaluation results

100%

Legal Document Semantic Search System

Pinecone vector store implementation

Criteria

Without context

With context

Modern Pinecone import

100%

ServerlessSpec import

100%

Instantiate Pinecone class

100%

ServerlessSpec with aws/us-east-1

100%

Idempotent index creation

100%

Batch upsert size

100%

Default cosine metric

100%

Default dimension 1536

100%

Metadata in search results

100%

Namespace support

100%

Filter support

100%

Result structure

100%

Without context: $0.3132 · 6m 27s · 15 turns · 113 in / 4,448 out tokens

With context: $0.4034 · 4m 26s · 16 turns · 647 in / 3,523 out tokens

100%

E-Commerce Product Search with PostgreSQL

pgvector HNSW indexing and hybrid search

Criteria

Without context

With context

Uses asyncpg

100%

Connection pool

100%

HNSW index type

100%

vector_cosine_ops operator

100%

HNSW m=16 parameter

100%

HNSW ef_construction=64

100%

Idempotent setup

100%

Cosine similarity operator

100%

Hybrid search method

100%

vector_weight parameter

100%

Metadata filter support

100%

Similarity score returned

100%

Without context: $0.4135 · 7m 3s · 14 turns · 105 in / 7,965 out tokens

With context: $0.4412 · 3m 50s · 14 turns · 692 in / 6,432 out tokens

100%

26%

Academic Research Paper Recommendation Engine

Qdrant with quantization and CrossEncoder reranking

Criteria

Without context

With context

QdrantClient import

66%

100%

Scalar quantization enabled

100%

INT8 quantization type

100%

Quantization quantile=0.99

100%

always_ram=True

100%

Idempotent collection creation

100%

CrossEncoder reranking

100%

CrossEncoder model name

100%

Over-fetch before rerank

100%

Rerank score assigned

50%

100%

Cosine distance metric

100%

PointStruct for upsert

100%

Without context: $1.0773 · 20m 44s · 33 turns · 257 in / 16,162 out tokens

With context: $0.6643 · 7m 41s · 22 turns · 169 in / 9,136 out tokens

Evaluated: about 18 hours ago
Agent: Claude Code

Table of Contents

Legal Document Semantic Search System E-Commerce Product Search with PostgreSQL Academic Research Paper Recommendation Engine

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.