全部能力

Scalable data processing for ML workloads. Streaming execution across CPU/GPU, supports Parquet/CSV/JSON/images. Integrates with Ray Train, PyTorch,…

使用《Scientific…》中的论文架构与写作策略原则，润色、重构学术文本，或将其翻译为符合《Nature》风格的英文。

Automatically generate **professional-grade** structured project Wiki from documentation, code, design files, and images. Use when: - User requests "generate wiki", "create docs", "create documentation" - User requests "update wiki", "rebuild wiki" - User requests "list plugins", "install plugin", "manage plugins" - Project needs automated documentation generation Features: - Smart project structure and tech stack analysis - **Deep code analysis** with semantic understanding - **Mermaid diagrams** for architecture, data flow, dependencies - **Cross-linked documentation** network - Incremental updates (only changed files) - Code blocks link to source files - Multi-language support (zh/en) - **Plugin system for extensions** For Chinese instructions, see references/SKILL.zh.md

State-space model with O(n) complexity vs Transformers' O(n²). 5× faster inference, million-token sequences, no KV cache. Selective SSM with hardware-aware…

RNN+Transformer hybrid with O(n) inference. Linear time, infinite context, no KV cache. Train like GPT (parallel), infer like RNN (sequential). Linux…

Implements and trains LLMs using Lightning AI's LitGPT with 20+ pretrained architectures (Llama, Gemma, Phi, Qwen, Mistral). Use when need clean model…

memory-search — an installable skill for AI agents, published by ruvnet/ruflo.

Orchestrate firmware rollouts with canary deployment and anomaly-gated advancement

Conduct technical research on technologies and architecture. Use when the user says they would like to do or produce a technical research report

research-synthesize — an installable skill for AI agents, published by ruvnet/ruflo.

Guides implementation of agent memory systems, compares production frameworks (Mem0, Zep/Graphiti, Letta, LangMem, Cognee), and designs persistence architectures for cross-session knowledge retention. Use when the user asks to "implement agent memory", "persist state across sessions", "build knowledge graph for agents", "track entities over time", "add long-term memory", "choose a memory framework", or mentions temporal knowledge graphs, vector stores, entity memory, adaptive memory, dynamic memory or memory benchmarks (LoCoMo, LongMemEval).

Run the SPARC Specification phase — gather requirements, define acceptance criteria, identify constraints, and store the spec in memory

Aggregate and display system metrics with anomaly detection for a time period

Extract entities and relations from source files to build a knowledge graph

Detect and classify candlestick patterns from ingested OHLCV data

Create and manage Cognitum Seed device fleets with firmware policies

Run the SPARC Refinement and Completion phases — review code, improve test coverage, validate against specification, and generate documentation

Create a new sequentially numbered database migration with up/down SQL files

Pathfinder traversal of the knowledge graph starting from a seed entity

Validate pending migrations for foreign key consistency, rollback safety, and best practices