Qdrant Search Quality

by qdrant6a03d0ce8f55No licenseListed Oct 8, 2026Updated Oct 8, 2026

Diagnoses and improves Qdrant search relevance. Use when someone reports 'search results are bad', 'wrong results', 'low precision', 'low recall', 'irrelevant matches', 'missing expected results', or asks 'how to improve search quality?', 'which embedding model?', 'should I use hybrid search?', 'how to combine keyword and vector search / fusion / RRF / prefetch?', 'should I use reranking?', 'relevance feedback?', 'how to measure retrieval quality?', 'build a golden set', 'ground truth dataset', or 'how to score recall@k?'. Also use when search quality degrades after quantization, model change, or data growth.

Instructions onlyAI & Agents Data & Analytics

AI-generated overview

Routes Qdrant search-quality complaints to guidance on diagnosis, hybrid search, reranking and relevance feedback.

What it does: This skill acts as a router for questions about poor Qdrant search relevance. It matches the user's reported symptom, such as low precision, low recall, missing expected results, or quality drops after quantization or a model change, to one of several guidance files and instructs the agent to read that file before answering. It also points to guidance on measuring retrieval quality, golden sets, recall@k, hybrid search with fusion or RRF, reranking, and relevance feedback. It produces routing decisions and answers drawn from the referenced guidance rather than from the routing page itself.
When to use it: Use it when someone reports bad, wrong, or irrelevant search results, missing expected matches, low precision or recall, or asks how to improve search quality. It also fits questions about embedding-model choice, hybrid keyword-plus-vector search, reranking, relevance feedback, and measuring retrieval quality with a golden set or recall@k.
Requirements: No scripts are shipped; it is instructions only. The agent needs file-reading tools (Read, Grep, Glob) and access to the referenced guidance files, which are not included in this skill folder.