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.

Quality

68%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./skills/vector-index-tuning/SKILL.md

The canonical home for this skill is vector-index-tuning in rmyndharis/antigravity-skills

Quality

Content

50%Scale 1-5

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

This skill provides a well-scoped but overly abstract guide to vector index tuning. Its main weakness is the near-complete lack of actionable, concrete guidance — no specific parameter values, no code examples, no benchmark commands. The skill essentially serves as a table of contents pointing to a resource file that doesn't exist in the bundle, making it hollow in practice.

Suggestions

Add concrete HNSW parameter examples (e.g., specific ef_construction, M values) and their trade-offs with executable code snippets for at least one vector database (e.g., pgvector, Pinecone, Qdrant).

Include a concrete quantization example showing before/after memory and recall impact, with actual code or configuration.

Add an explicit feedback loop in the workflow: 'If recall drops below target after parameter change, revert parameters and try smaller increments' with specific validation commands.

Either provide the referenced `resources/implementation-playbook.md` bundle file or inline the most critical patterns and checklists directly in the skill.

Dimension	Reasoning	Score
Conciseness	The content is fairly lean and doesn't over-explain concepts Claude already knows. The 'Use this skill when' and 'Do not use this skill when' sections add some bulk but are reasonable for scoping. Minor trimming possible (e.g., the resource is mentioned twice).	4 / 5
Actionability	The instructions are high-level directives ('gather workload targets', 'choose an index type', 'benchmark parameter sweeps') without any concrete code, commands, specific parameter values, or examples. There are no executable snippets, no example HNSW parameter configurations, no quantization code, and no benchmark commands. The real content is deferred to a resource file that doesn't exist in the bundle.	2 / 5
Workflow Clarity	There is a 4-step sequence with a validation step (step 4: validate on staging before production), and the safety section mentions rollback plans and recall regression tracking. However, the steps are vague, there are no explicit validation commands or feedback loops (e.g., what to do if recall drops during parameter sweep), and the destructive operation of reindexing lacks concrete rollback procedures.	3 / 5
Progressive Disclosure	The skill references `resources/implementation-playbook.md` for detailed content, which is a good structural choice. However, the bundle file doesn't exist (no bundle files provided), so the reference is unverifiable. The reference is mentioned twice (in Instructions and Resources), and there's only one external file referenced — the skill itself is too thin without it, suggesting too much content was deferred.	3 / 5
	Total	12 / 20 Passed

Description

87%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.

This is a strong, well-crafted skill description that clearly defines a specific niche in vector search optimization. It effectively uses the 'Use when' pattern with concrete trigger scenarios and domain-specific terminology. Minor improvements could include broader trigger term coverage with synonyms and related technologies.

Dimension	Reasoning	Score
Specificity	Lists several specific actions: optimizing for latency/recall/memory, tuning HNSW parameters, selecting quantization strategies, and scaling vector search infrastructure. These are concrete but could be more comprehensive (e.g., mentioning specific quantization types, benchmarking, or specific vector databases).	4 / 5
Completeness	Clearly answers both 'what' (optimize vector index performance for latency, recall, and memory) and 'when' (tuning HNSW parameters, selecting quantization strategies, or scaling vector search infrastructure) with explicit 'Use when' clause and concrete trigger scenarios.	5 / 5
Trigger Term Quality	Includes strong domain-specific trigger terms like 'HNSW', 'quantization', 'vector search', 'vector index', 'latency', 'recall'. Missing some natural user phrases like 'ANN search', 'approximate nearest neighbor', 'embedding search', 'similarity search', or specific tool names like 'FAISS', 'Pinecone', 'Milvus'.	4 / 5
Distinctiveness Conflict Risk	Highly distinctive niche focusing specifically on vector index performance tuning. The combination of HNSW, quantization, and vector search infrastructure creates a very clear and unique domain that is unlikely to conflict with other skills.	5 / 5
	Total	18 / 20 Passed

Validation

90%

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 — 10 / 11 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	10 / 11 Passed

Repository: administrakt0r/AI-Agents-Safe-Coding-Skills
Path: skills/vector-index-tuning/SKILL.md
Commit: 3b73fbb

Reviewed: ~16 hours 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.