g14wxz/hybrid-search-rrf-pattern
Fuses semantic vector search with lexical full-text search using Reciprocal Rank Fusion in a PL/pgSQL RPC.
100
100%
Does it follow best practices?
Impact
Pending
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 an excellent skill description that clearly defines a narrow, specific capability with concrete technical details. It includes an explicit 'Use when' clause with multiple natural trigger terms covering both technical jargon and conceptual descriptions. The description is concise yet comprehensive, making it easy for Claude to select this skill precisely when needed.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Lists multiple specific concrete actions: creates PL/pgSQL RPC, implements Reciprocal Rank Fusion with explicit formula (score = 1/(k+rank)), fuses semantic pgvector results with full-text tsvector results. Very concrete and technical. | 3 / 3 |
Completeness | Clearly answers both 'what' (creates PL/pgSQL RPC implementing RRF to fuse semantic and full-text results) and 'when' (explicit 'Use when' clause listing hybrid search, RRF search, semantic plus keyword search, SKU queries, vector and full-text fusion). | 3 / 3 |
Trigger Term Quality | Excellent coverage of natural terms users would say: 'hybrid search', 'RRF search', 'semantic plus keyword search', 'exact SKU and conceptual queries', 'vector and full-text fusion', 'pgvector', 'tsvector', 'PL/pgSQL'. Covers both technical and conceptual variations. | 3 / 3 |
Distinctiveness Conflict Risk | Highly distinctive niche combining PL/pgSQL, RRF, pgvector, and tsvector. Very unlikely to conflict with other skills due to the specific technology stack and algorithm mentioned. | 3 / 3 |
Total | 12 / 12 Passed |
Implementation
100%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is an excellent skill that provides a complete, executable hybrid search RPC with clear validation and verification phases. It respects Claude's intelligence by avoiding unnecessary explanations while providing all the concrete details needed for deployment. The HALT conditions and verification criteria make the workflow robust against common failure modes.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The content is lean and efficient. Every section serves a purpose — preconditions, schema validation, the SQL function, verification, and tuning notes. Comments within the SQL are minimal and useful. No unnecessary explanations of what RRF is conceptually or how pgvector works at a basic level. | 3 / 3 |
Actionability | Provides a complete, executable SQL function with clear placeholders to substitute. Verification steps include specific SQL commands (EXPLAIN ANALYZE) and concrete success criteria. The schema validation phase gives exact queries to run. | 3 / 3 |
Workflow Clarity | Three clearly sequenced phases (Validate → Deploy → Verify) with explicit HALT conditions at validation and verification stages. The verification phase includes a feedback loop checking index usage and score correctness, with a clear stop condition if sequential scans appear. | 3 / 3 |
Progressive Disclosure | Well-structured with clear sections. References the pgvector-hnsw-index-selection skill as a one-level-deep dependency. Tuning notes are appropriately separated at the end rather than cluttering the main workflow. The content length is appropriate for inline presentation without needing external files. | 3 / 3 |
Total | 12 / 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.
Reviewed
Table of Contents