Rewrite ambiguous queries by clarifying terms and expanding abbreviations.
npx skills add https://github.com/edangx100/adaptive_rag --skill rewriting
Please help me install this Agent Skill. Skill: rewriting Source: https://github.com/edangx100/adaptive_rag/tree/main/.claude/skills/rewriting Command: npx skills add https://github.com/edangx100/adaptive_rag --skill rewriting
This Skill eliminates the frustration of failed searches by automatically improving unclear or incomplete queries for better results.
Rewrite the query "fast computer" to improve search results.
High-intent search queries and answers about installing and using this skill.
Query rewriting automatically enhances vague or incomplete searches by clarifying terms, expanding abbreviations, and adding context. When a search fails to find relevant documents, the Skill rewrites the query to improve retrieval performance and surface better results.
This Skill rewrites ambiguous queries by incorporating synonyms, expanding context, and clarifying vague terminology. For example, 'won't work' becomes 'troubleshooting device not functioning properly,' enabling search systems to locate relevant documents.
Query rewriting works as an adaptive retry mechanism when initial searches fail to find relevant documents. It's invoked automatically when no relevant results are found, improving query clarity without requiring manual intervention or additional user input.
Yes. Query rewriting integrates into RAG pipelines as a retry strategy, enhancing search performance when document retrieval fails. It clarifies queries to increase the likelihood of finding relevant documents for generation tasks.
Yes. The Skill expands abbreviations, clarifies vague terms, and adds context to technical or informal queries. This makes it effective for searches containing domain-specific language, acronyms, or colloquial phrasing that standard retrieval might miss.
patrykkopycinskiSet up a RAG pipeline with Elasticsearch for vector retrieval.
NotysotyDesign hybrid retrieval pipelines combining dense vector search with BM25 keyword matching.
ionmidoriCombine vector and keyword search results using Reciprocal Rank Fusion.
enoch-robinsonDesign role prompts, few-shot templates, and system prompts for LLMs.
