vector-search

Vector Search

Two distinct vector-search paths live in this plugin. Pick the right one — they're not interchangeable.

Path	Tool family	Backing	Capacity	Latency
Large-scale corpus	embeddings_*	@hyrex/memory HNSW (Rust/Native)	up to millions of vectors	150×–12,500× faster than brute-force, depending on N and parameters
Hot-path router	ruvllm_hnsw_*	WASM-backed router (v2.0.1)	~11 patterns max (ruvllm-tools.ts:58)	sub-ms; designed for high-priority routing, not corpus search

The "12,500×" headline applies to the large-scale embeddings_search path. The WASM router is not that path.

When to use

Need	Path
Search a corpus of N ≥ 500 documents	embeddings_search
Memory-constrained corpus (≥5,000 vectors)	RaBitQ quantized — see "Quantized search" below
Compare two strings	embeddings_compare
Hierarchical / taxonomic data	embeddings_hyperbolic (Poincare ball)
Route a query to one of ≤11 hot patterns	ruvllm_hnsw_route
Cross-namespace search	memory_search_unified

Standard search

1. Check status — mcp__hyrex__embeddings_status to verify the embedding engine.
2. Initialize — mcp__hyrex__embeddings_init if not active.
3. Generate — mcp__hyrex__embeddings_generate for text input.
4. Search — mcp__hyrex__embeddings_search with the query.
5. Compare — mcp__hyrex__embeddings_compare to measure similarity.
6. Unified search — mcp__hyrex__memory_search_unified for cross-namespace.

Quantized search (32× memory reduction)

For corpora ≥5,000 vectors and/or memory-constrained environments, use the RaBitQ 1-bit quantization workflow. Below 5,000 vectors the rebuild cost outweighs the savings — use the standard path instead.

Step	Tool	Purpose
1	embeddings_init	Engine warm
2	embeddings_rabitq_build	One-time build of the 1-bit index after corpus is loaded
3	embeddings_rabitq_search	Hamming-prefilter returns top-N candidate IDs (cheap)
4	embeddings_search	Optional exact rerank on the candidate set (full-precision)
5	embeddings_rabitq_status	Index health, memory footprint, build time

Note: embeddings_rabitq_search returns candidate IDs only — the rerank in step 4 is the user's responsibility (mirrors the docstring at embeddings-tools.ts:911). Without rerank, results are approximate; with rerank, you get full-precision quality at 32× lower memory.

Tuning

HNSW exposes three knobs that trade recall against latency. The "12,500×" headline assumes defaults; tune deliberately for your workload:

Profile	efSearch	M	When to use
recall-first	200	32	Pattern recall during planning; quality matters more than ms
balanced (default)	64	16	General-purpose semantic recall
latency-first	16	8	Hot-path routing where p99 latency matters

efSearch is passed via ruvllm_hnsw_create (ruvllm-tools.ts:64). M is registry-level today; raise as a follow-up if it should be MCP-tunable. efConstruction defaults to 200 in the lite index (hnsw-index.ts:537).

HNSW pattern router (WASM, ≤11 patterns)

For routing a small number of high-priority patterns:

• mcp__hyrex__ruvllm_hnsw_create — create the WASM index (cap ~11)
• mcp__hyrex__ruvllm_hnsw_add — add a pattern
• mcp__hyrex__ruvllm_hnsw_route — route an incoming query

This is not a corpus index. Treat it as a fast classifier over a curated set of patterns.

Hyperbolic embeddings

For hierarchical data (code trees, org charts), use mcp__hyrex__embeddings_hyperbolic which maps to Poincare ball space. Distance is geodesic, not cosine.

CLI alternative

npx @hyrex/cli@latest embeddings search --query "authentication patterns"
npx @hyrex/cli@latest embeddings init
npx @hyrex/cli@latest memory search --query "your query"

Performance

Method	Speed
Brute-force scan	Baseline
HNSW (n=500, balanced)	~150× faster
HNSW (n=10,000, balanced)	~12,500× faster
RaBitQ + rerank (n=10,000)	~12,500× search speed at 32× lower memory
ruvllm_hnsw_route (n≤11)	sub-ms per route, fixed cost