hybrid-search-implementation
Guides hybrid search implementation: vector embeddings, keyword search, fusion algorithms, caching strategies. Use when implementing search, recommendation systems, or discovery features.
Découvrez les skills tools academic, comparez leurs dépôts et consultez leurs instructions.
Guides hybrid search implementation: vector embeddings, keyword search, fusion algorithms, caching strategies. Use when implementing search, recommendation systems, or discovery features.
Use when performing deep technical research, evaluating libraries, comparing options, or investigating APIs. Covers documentation lookup, issue investigation, version compatibility analysis, and synthesis of fi...
Enforces iterative web research for coding and DevOps questions to prevent assumptions about libraries, frameworks, APIs, versions, and configurations. Use this skill when the user asks about programming langua...
Generate research material on a given topic, placing content in the in-research folder. Focuses on content aggregation from different sources and outlining what should be covered in a future article.
Research subject matter, align to educational standards, map prerequisites, and recommend learning theories for curriculum design. Use when starting curriculum development, exploring new topics, or needing stan...
Assist with grading by applying rubrics to student work, generating criterion-level feedback, and maintaining consistency across submissions. Use when grading assignments, applying rubrics, or providing feedbac...
Author high-quality assessment items (questions, prompts, tasks) aligned to learning objectives with answer keys and rubrics. Use when creating test questions, writing assessment items, or building item banks....
Design peer assessment systems including rubrics, calibration activities, review protocols, and feedback quality guidelines. Use for peer learning. Activates on "peer review", "peer assessment", "peer feedback"...
Detects plagiarism and validates content originality for educational materials.
Remove signs of AI-generated writing from text. Use when editing or reviewing text to make it sound more natural and human-written. Based on Wikipedia's comprehensive "Signs of AI writing" guide. Detects and fi...
Remove signs of AI-generated writing from text. Use when editing or reviewing text to make it sound more natural and human-written. Based on Wikipedia's comprehensive "Signs of AI writing" guide. Detects and fi...
Irreducible Semantic Density Analysis - analyzes text compressibility by decomposing into five strata (structure, retrievable, derived, curatorial, novel). Use when asked to analyze text density, measure origin...
Aggregates user reviews from Amazon/Reddit and expert reviews from specialized sites with sentiment analysis. Use when user asks to 'get reviews', 'what do users say', 'aggregate reviews', 'user feedback', 'exp...
발굴조사 보고서 고찰(考察) 섹션 초안 작성
Use xAI's Grok model with agentic tool calling for X (Twitter) search, web search, code execution, and real-time data access. Invoke when user needs Twitter/X insights, current events, alternative perspectives,...
국가유산청 발굴조사 보고서 규정 준수 체크리스트 생성
Hybrid search (embedding + BM25) for retrieving relevant clinical note passages. Use for finding source evidence to support claims in summaries and recommendations.
발굴조사 자료(논문/보고서/주변유적) 수집 및 메타데이터 정규화
Contextual Retrieval implementation for RAG - chunks clinical notes with LLM-generated context prepended to each chunk before embedding. Improves citation accuracy by 49% per Anthropic research.
Ensure the library has collected all publications by faculty in a department - a collection completeness requirement
literature-review for evidence-based learning research and evaluation.
Standard collection development workflow for discovering, evaluating, and acquiring scholarly materials for the library collection
Extract and validate citations with Jaccard overlap calculation. Ensures all claims have verifiable source evidence with ≥0.7 alignment threshold.
data-collection for evidence-based learning research and evaluation.
Le classement utilise les étoiles du dépôt GitHub indexé. Plusieurs skills peuvent partager ce nombre : ce n’est ni un nombre d’installations ni une note de qualité.