What is search intent in Bluera Knowledge?

Search intent specifies the goal of your query (e.g., finding an implementation vs. a definition) to help the engine rank the most relevant files higher based on your specific needs.

How can I reduce token usage during large searches?

Use the 'minimal' detail level for initial discovery to get summaries, then use 'get_full_context' only for the few results with the highest relevance scores.

When should I use the 'contextual' detail level?

Choose 'contextual' when you need to see function signatures, types, and imports to understand an interface without needing the full internal implementation logic.

Can I search multiple repositories simultaneously?

Yes, by using the 'stores' parameter, you can target specific indexed libraries or search across all available stores for a broader comparative perspective.

What is a good relevance score for a search result?

Scores between 0.9 and 1.0 are excellent matches, while 0.7 to 0.9 are very likely relevant. Results below 0.5 are typically weak matches and can often be ignored.

Bluera Search Optimization

Name: Bluera Search Optimization
Author: blueraai

byblueraai

0•

学習とドキュメンテーション

Optimizes Bluera Knowledge search queries by implementing intent-based ranking and token-efficient retrieval strategies.

The Search Optimization skill provides a comprehensive framework for maximizing the effectiveness of the Bluera Knowledge search tool within Claude Code. It guides users through advanced parameter management, including the selection of specific search intents like implementation discovery, pattern finding, and API definitions. By employing a progressive context strategy—starting with minimal summaries and escalating to full code only when necessary—this skill helps developers navigate vast codebases and documentation sets while reducing token consumption by up to 70%. It is an essential utility for developers who need high-precision results from their local knowledge stores without overwhelming the LLM's context window.

主な機能

01Intent-based ranking for implementation, usage, and documentation queries

02Balanced discovery strategies for both wide-net and high-precision searches

03Progressive detail levels to minimize token overhead and context usage

04Multi-store filtering for targeted search across specific repositories

05Relevance score evaluation for data-driven result selection

060 GitHub stars

ユースケース

01Exploring large unfamiliar codebases without exceeding context limits

02Performing comparative analysis of implementations across different frameworks

03Extracting specific API signatures and type definitions quickly

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add blueraai/bluera-knowledge search-optimization

For use in Claude.ai and ChatGPT

Download Skill