databricks-vector-search
Patterns for Databricks Vector Search: create endpoints and indexes, query with filters, manage embeddings. Use when building RAG applications, semantic search, or similarity matching. Covers both storage-optimized and standard endpoints.
71
86%
Does it follow best practices?
Impact
—
No eval scenarios have been run
Passed
No known issues
Quality
Discovery
100%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 skill description that clearly communicates specific capabilities, includes natural trigger terms, and explicitly states both what the skill does and when to use it. The Databricks Vector Search focus provides a clear niche that minimizes conflict risk with other skills. The description is concise yet comprehensive.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Lists multiple specific concrete actions: 'create endpoints and indexes, query with filters, manage embeddings' and further specifies 'storage-optimized and standard endpoints'. These are concrete, actionable capabilities. | 3 / 3 |
Completeness | Clearly answers both what ('create endpoints and indexes, query with filters, manage embeddings') and when ('Use when building RAG applications, semantic search, or similarity matching'). The 'Use when...' clause is explicit and well-defined. | 3 / 3 |
Trigger Term Quality | Includes strong natural keywords users would say: 'Databricks Vector Search', 'RAG applications', 'semantic search', 'similarity matching', 'embeddings', 'endpoints', 'indexes'. These cover the main terms a user working in this domain would naturally use. | 3 / 3 |
Distinctiveness Conflict Risk | Highly distinctive with 'Databricks Vector Search' as a clear niche. The combination of Databricks-specific terminology with vector search concepts makes it very unlikely to conflict with other skills. | 3 / 3 |
Total | 12 / 12 Passed |
Implementation
72%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a comprehensive, well-organized skill with excellent actionability — every pattern includes executable Python code and CLI equivalents. The progressive disclosure is well-handled with clear references to supporting files. The main weaknesses are some redundancy between SDK examples and MCP tool examples, and the lack of explicit multi-step workflow sequencing with validation checkpoints (e.g., checking endpoint readiness before index creation).
Suggestions
Add an explicit end-to-end workflow section with numbered steps and validation checkpoints: create endpoint → poll until READY → create index → poll until ONLINE → query, with status-checking code at each step.
Reduce redundancy by consolidating the MCP Tools section — either reference the SDK examples above or remove the duplicate SDK-style code blocks, keeping only the MCP-specific function signatures and parameters.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is generally well-structured but is quite long (~300+ lines) with some redundancy. The MCP Tools section largely duplicates the SDK patterns shown in Quick Start and Common Patterns. The overview tables and 'When to Use' section explain things Claude could infer. However, the comparison tables and filter syntax differences are genuinely useful non-obvious information. | 2 / 3 |
Actionability | The skill provides fully executable, copy-paste-ready Python code for every operation: creating endpoints, creating all three index types, querying with text/vector/hybrid, filtering with both syntaxes, upserting, deleting, scanning, and syncing. CLI commands are also complete and specific. | 3 / 3 |
Workflow Clarity | While individual operations are clear, there's no explicit end-to-end workflow with validation checkpoints. Endpoint creation is noted as asynchronous but there's no explicit 'wait for ready' step before creating indexes. Index creation → sync → query flow lacks explicit sequencing with status checks. The reference to end-to-end-rag.md suggests this exists elsewhere, but the SKILL.md itself doesn't provide a validated multi-step workflow. | 2 / 3 |
Progressive Disclosure | The Reference Files table clearly signals four one-level-deep references (index-types.md, end-to-end-rag.md, search-modes.md, troubleshooting-and-operations.md) with descriptions. The SKILL.md serves as a comprehensive overview with quick-start patterns while deferring detailed walkthroughs and decision guides to separate files. Related Skills section also provides clear cross-references. | 3 / 3 |
Total | 10 / 12 Passed |
Validation
100%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 11 / 11 Passed
Validation for skill structure
No warnings or errors.
- Repository
- databricks-solutions/ai-dev-kit
- Commit
93cb4e3
- Reviewed
Table of Contents
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.