vector-index-tuning

Optimize vector index performance for latency, recall, and memory. Use when tuning HNSW parameters, selecting quantization strategies, or scaling vector search infrastructure.

1.56x

Quality

71%

Does it follow best practices?

Impact

100%

1.56x

Average score across 3 eval scenarios

Securityby

Passed

No known issues

Optimize this skill with Tessl

npx tessl skill review --optimize ./tests/ext_conformance/artifacts/agents-wshobson/llm-application-dev/skills/vector-index-tuning/SKILL.md

Evaluation results

100%

10%

HNSW Index Performance Benchmarking

HNSW parameter benchmarking

Criteria

Without context

With context

hnswlib used

100%

Cosine space

100%

Set-intersection recall

100%

Graph edge memory

100%

Four metrics recorded

100%

M=32 in sweep

100%

efConstruction ≥200 tested

100%

efSearch sweep

100%

benchmark_results.json

100%

requirements.txt with hnswlib

100%

37%

Vector Compression Library

Quantization strategies and memory estimation

Criteria

Without context

With context

INT8 scale formula

100%

Clip and round to uint8

100%

Params dict contents

30%

100%

KMeans for PQ

100%

PQ divisibility assertion

100%

PQ default parameters

100%

Binary sign threshold

100%

Bit packing with bitwise OR

100%

Bytes-per-dim constants

100%

HNSW graph overhead

100%

60%

Qdrant Collection Configuration Profiles

Qdrant collection configuration profiles

Criteria

Without context

With context

Recall profile HNSW

50%

100%

Speed profile HNSW

50%

100%

Memory profile HNSW

100%

Speed quantization

100%

Memory quantization

100%

Memory optimizer config

100%

Speed optimizer config

100%

Search params for ≥0.99 recall

100%

Search params for ≥0.95 recall

50%

100%

Balanced profile HNSW

50%

100%

Repository: Dicklesworthstone/pi_agent_rust
Commit: 47823e3

Evaluated: about 1 month ago
Agent: Claude Code
Model: Claude Sonnet 4.6

Table of Contents

HNSW Index Performance Benchmarking Vector Compression Library Qdrant Collection Configuration Profiles

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.