vector-index-tuning — quality + safety report

In the Skillier index (wshobson-agents__vector-index-tuning) · scanned 2026-06-03 · engine: builtin+triage

A
Quality
98/100
Safety

✓ Clean — no heuristic safety flags surfaced.

Heuristic flags from the builtin scanner, which is known to over-flag (it trips on legitimate env-reading integrations, security skills, and library .eval calls). This is NOT an authoritative malicious verdict — re-scan with SkillSpector for the authoritative result. Run the authoritative scan →

Skillproof quality grade A

📇 This skill is in the Skillier index (curated · deduped · quality-filtered). Install Skillier to route & load it into your AI client.

Quality notes

No explicit output format / contract
low · quality · body
→ State the expected output format (structure, sections, or schema).

About this skill

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

📄 Read the SKILL.md
---
name: vector-index-tuning
description: Optimize vector index performance for latency, recall, and memory. Use when tuning HNSW parameters, selecting quantization strategies, or scaling vector search infrastructure.
---

# Vector Index Tuning

Guide to optimizing vector indexes for production performance.

## When to Use This Skill

- Tuning HNSW parameters
- Implementing quantization
- Optimizing memory usage
- Reducing search latency
- Balancing recall vs speed
- Scaling to billions of vectors

## Core Concepts

### 1. Index Type Selection

```
Data Size           Recommended Index
────────────────────────────────────────
< 10K vectors  →    Flat (exact search)
10K - 1M       →    HNSW
1M - 100M      →    HNSW + Quantization
> 100M         →    IVF + PQ or DiskANN
```

### 2. HNSW Parameters

| Parameter          | Default | Effect                                               |
| ------------------ | ------- | ---------------------------------------------------- |
| **M**              | 16      | Connections per node, ↑ = better recall, more memory |
| **efConstruction** | 100     | Build quality, ↑ = better index, slower build        |
| **efSearch**       | 50      | Search quality, ↑ = better recall, slower search     |

### 3. Quantization Types

```
Full Precision (FP32): 4 bytes × dimensions
Half Precision (FP16): 2 bytes × dimensions
INT8 Scalar:           1 byte × dimensions
Product Quantization:  ~32-64 bytes total
Binary:                dimensions/8 bytes
```

## 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

- **Benchmark with real queries** - Synthetic may not represent production
- **Monitor recall continuously** - Can degrade with data drift
- **Start with defaults** - Tune only when needed
- **Use quantization** - Significant memory savings
- **Consider tiered storage** - Hot/cold data separation

### Don'ts

- **Don't over-optimize early** - Profile first
- **Don't ignore build time** - Index updates have cost
- **Don't forget reindexing** - Plan for maintenance
- **Don't skip warming** - Cold indexes are slow
Scan or optimize your own skill →

Want a live grade + an embeddable README badge? Run your skill through the free scanner.

Graded independently by Skillproof — nothing to sell the author. Quality is mechanical + corpus-grounded; safety flags are heuristic (builtin+triage), not a malicious verdict.