Universal Runtime best practices for PyTorch inference, Transformers models, and FastAPI serving. Covers device management, model loading, memory optimization, and performance tuning.
Expert security analyst using ALL MCP servers. Uses Memory for known vulnerabilities, UltraThink for threat modeling, search MCPs for CVEs, and MongoDB for audit logs.
Vaporwave + glassomorphic UI designer for photo/memory apps. Masters SwiftUI Material effects, neon pastels, frosted glass blur, retro-futuristic design. Expert in 2025 UI trends (glassmorphism,...
Activation-aware weight quantization for 4-bit LLM compression with 3x speedup and minimal accuracy loss. Use when deploying large models (7B-70B) on limited GPU memory, when you need faster...
Create flashcards for spaced repetition learning, compatible with Anki import. Supports basic cards, cloze deletions, and mnemonic aids. Use when creating study flashcards, Anki decks, vocabulary...
Implement RxJS observables, apply operators, fix memory leaks with unsubscribe patterns, handle errors, create subjects, and build reactive data pipelines in Angular applications.
Design and evaluate context compression strategies for long-running agent sessions. Use when agents exhaust memory, need to summarize conversation history, or when optimizing tokens-per-task...
Embedding and vector retrieval expert for semantic searchUse when "vector search, embeddings, semantic search, qdrant, pgvector, similarity search, reranking, hybrid retrieval, embeddings,...
Expert in ALL computational collage composition: photo mosaics, grid layouts, scrapbook/journal styles, magazine editorial, vision boards, mood boards, social media collages, memory walls,...
Develop ICP canister backends using Motoko. Covers actors, types, async patterns, stable memory, upgrades, access control, and error handling. Use when writing Motoko code, creating canisters,...
Performance optimization mindset - knowing when to optimize, how to measure, where bottlenecks hide, and when "fast enough" is the right answerUse when "slow, performance, optimize, profiling,...
Language-independent tokenizer treating text as raw Unicode. Supports BPE and Unigram algorithms. Fast (50k sentences/sec), lightweight (6MB memory), deterministic vocabulary. Used by T5, ALBERT,...
Language-independent tokenizer treating text as raw Unicode. Supports BPE and Unigram algorithms. Fast (50k sentences/sec), lightweight (6MB memory), deterministic vocabulary. Used by T5, ALBERT,...
Data pipeline specialist for ETL design, data quality, CDC patterns, and batch/stream processingUse when "data pipeline, etl, cdc, data quality, batch processing, stream processing, data...
Fast in-process analytical database for SQL queries on DataFrames, CSV, Parquet, JSON files, and more. Use when user wants to perform SQL analytics on data files or Python DataFrames (pandas,...
Memory sidecar for agent work: recall before tasks, record learnings after tasks, review recommendations, optional backport bundles.
Use when about to ask the user a factual question, propose a solution, diagnose an error, or choose between approaches. Triggers on: 'Do you have X installed?', 'What version?', 'Is X...
Expert guidance for fast fine-tuning with Unsloth - 2-5x faster training, 50-80% less memory, LoRA/QLoRA optimization
Expert guidance for fast fine-tuning with Unsloth - 2-5x faster training, 50-80% less memory, LoRA/QLoRA optimization
Write stable learnings, decisions, and patterns to durable storage like CLAUDE.md or knowledge files. Use when saving project decisions, recording patterns, or updating long-term memory.