neo4j-vector-index-skill

Create and manage Neo4j vector indexes, run vector similarity search (ANN/kNN), store embeddings on nodes or relationships, use SEARCH clause (Neo4j 2026.01+, preferred) or db.index.vector.queryNodes() procedure (deprecated 2026.04, still works on 2025.x), configure HNSW and quantization options, pick similarity function and embedding provider dimensions, and batch-update embeddings. Use when tasks involve CREATE VECTOR INDEX, vector.dimensions, cosine/euclidean search, embedding ingestion pipelines, semantic or structural nearest-neighbor lookup, or hybrid search (vector + fulltext, multiple vector sources, or graph-derived scores). Does NOT handle GraphRAG retrieval_query graph traversal — use neo4j-graphrag-skill. Does NOT handle fulltext-only/keyword-only search — use neo4j-cypher-skill. Does NOT compute GDS graph embeddings (FastRP, Node2Vec) — use neo4j-gds-skill.

0.96x

Quality

90%

Does it follow best practices?

Impact

63%

0.96x

Average score across 3 eval scenarios

Securityby

Passed

No findings from the security scan

Quality

Content

81%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body is a highly actionable, well-sequenced workflow with strong validation checkpoints and an appropriately offloaded hybrid-search reference file. Its main weakness is redundancy — similarity-function and version-preference guidance is repeated across multiple sections — which costs token budget without adding clarity.

Suggestions

Consolidate the similarity-function guidance: Step 1's table plus the wrong-similarity pitfall appear again verbatim in 'Similarity Function — Extended Guidance' and the Gotchas table — collapse into a single authoritative section and cross-reference it.

Reduce repetition of the SEARCH-vs-procedure version note (it recurs in Pre-flight, Step 1, Step 4, Checklist, and ai.text.embed sections) — state the version rule once and reference it.

Consider offloading the Embedding Provider Quick-Reference and Vector Functions tables into a references file (linked from References) to slim the main body, since they are lookup material rather than core workflow.

Dimension	Reasoning	Score
Conciseness	Mostly efficient reference-style content (config tables, code, error tables) but notable redundancy: the similarity-function guidance is restated across Step 1, the 'Similarity Function — Extended Guidance' section, and the Gotchas table, and the SEARCH-vs-procedure version note is repeated several places, so it could be tightened beyond minor trims.	3 / 5
Actionability	Abundant copy-paste-ready, executable Cypher and Python (index creation, ONLINE polling, UNWIND batch ingest, SEARCH/post-filter queries) covering the common cases, matching the anchor-5 example.	5 / 5
Workflow Clarity	Clear sequenced workflow (Pre-flight → Steps 1–6) with explicit validation checkpoints (poll to ONLINE, dimension assertion before ingest), error-recovery guidance, and a final Checklist — feedback loops appropriate for this batch/index-rebuild skill.	5 / 5
Progressive Disclosure	Good structure with clearly signaled one-level-deep reference (references/hybrid-search.md) loaded in Step 6 and listed in References, with hybrid-search bulk appropriately offloaded; however the body itself is large (~506 lines) and retains several sections (embedding-provider table, vector functions, extended guidance) inline that could arguably be referenced, keeping it just below a clean 5.	4 / 5
	Total	17 / 20 Passed

Description

100%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description is comprehensive and well-structured, explicitly covering both capabilities and trigger conditions while cleanly distinguishing the skill from three adjacent Neo4j skills via explicit negative boundary clauses. It assumes third-person voice and avoids vague fluff, with rich synonym coverage of trigger terms.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions — 'Create and manage Neo4j vector indexes, run vector similarity search (ANN/kNN), store embeddings... configure HNSW and quantization options, pick similarity function... batch-update embeddings' — with comprehensive coverage, matching the anchor-5 example.	5 / 5
Completeness	Explicitly answers both 'what' (the action list) and 'when' via 'Use when tasks involve...' with concrete trigger phrases, plus explicit boundary exclusions, matching the anchor-5 example.	5 / 5
Trigger Term Quality	Comprehensive natural/technical trigger coverage with synonyms — 'CREATE VECTOR INDEX, vector.dimensions, cosine/euclidean search, embedding ingestion pipelines, semantic or structural nearest-neighbor lookup... hybrid search' — including paired variants (cosine/euclidean, ANN/kNN), matching the comprehensive-synonyms anchor.	5 / 5
Distinctiveness Conflict Risk	Clear Neo4j-vector niche with explicit negative boundaries routing three sibling scopes elsewhere ('Does NOT handle GraphRAG... use neo4j-graphrag-skill', 'Does NOT handle fulltext-only... use neo4j-cypher-skill', 'Does NOT compute GDS graph embeddings... use neo4j-gds-skill'), giving minimal conflict risk.	5 / 5
	Total	20 / 20 Passed

Validation

81%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 13 / 16 Passed

Validation for skill structure

Criteria	Description	Result
skill_md_line_count	SKILL.md is long (506 lines); consider splitting into references/ and linking	Warning
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning
relative_links	Relative link issues: 1 suspicious	Warning

	Total	13 / 16 Passed

Repository: neo4j-contrib/neo4j-skills
Path: neo4j-vector-index-skill/SKILL.md
Commit: a9a5e78

Reviewed: 6 days ago

Table of Contents

Discovery Implementation Validation

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.