MLOps
模型部署、评估与运维
business-model-preferences
phrazzld
Pricing philosophy and business model constraints. Auto-invoke when: evaluating pricing, checkout flows, subscription logic, tier structures.
transformers-llm
levy-n
Implements Transformer models and LLM workflows. Covers attention mechanism, BERT fine-tuning, HuggingFace Transformers library (Tokenizer, Trainer, Pipeline), and LLM ecosystem (GPT, Claude, Gemini, Ollama). Use when fine-tuning language models, using HuggingFace, calling LLM APIs, or when user mentions 'transformer', 'attention', 'BERT', 'HuggingFace', 'tokenizer', 'fine-tuning', 'LLM', 'GPT', 'Claude', 'Gemini', 'prompt engineering', 'zero-shot', or 'few-shot learning'.
deepchem
hxk622
Molecular ML with diverse featurizers and pre-built datasets. Use for property prediction (ADMET, toxicity) with traditional ML or GNNs when you want extensive featurization options and MoleculeNet benchmarks. Best for quick experiments with pre-trained models, diverse molecular representations. For graph-first PyTorch workflows use torchdrug; for benchmark datasets use pytdc.
scvi-tools
hxk622
Deep generative models for single-cell omics. Use when you need probabilistic batch correction (scVI), transfer learning, differential expression with uncertainty, or multi-modal integration (TOTALVI, MultiVI). Best for advanced modeling, batch effects, multimodal data. For standard analysis pipelines use scanpy.
debug:scikit-learn
SnakeO
Debug Scikit-learn issues systematically. Use when encountering model errors like NotFittedError, shape mismatches between train and test data, NaN/infinity value errors, pipeline configuration issues, convergence warnings from optimizers, cross-validation failures due to class imbalance, data leakage causing suspiciously high scores, or preprocessing errors with ColumnTransformer and feature alignment.
tavus-replica
Tavus-Engineering
Create and manage Tavus replicas (AI digital twins). Use when training custom replicas from video, listing stock replicas, or managing replica assets. Covers training video requirements, consent statements, and the Phoenix-3 model.
tavus-overview
Tavus-Engineering
Overview of Tavus, the AI research lab pioneering human computing. Use when you need context about what Tavus is, their mission, core concepts like CVI and Human Computing, the model stack (Phoenix, Raven, Sparrow), or links to docs/platform/resources.
debug:tensorflow
SnakeO
Debug TensorFlow and Keras issues systematically. This skill helps diagnose and resolve machine learning problems including tensor shape mismatches, GPU/CUDA detection failures, out-of-memory errors, NaN/Inf values in loss functions, vanishing/exploding gradients, SavedModel loading errors, and data pipeline bottlenecks. Provides tf.debugging assertions, TensorBoard profiling, eager execution debugging, and version compatibility guidance.
advanced-evaluation
xfstudio
This skill should be used when the user asks to "implement LLM-as-judge", "compare model outputs", "create evaluation rubrics", "mitigate evaluation bias", or mentions direct scoring, pairwise comparison, position bias, evaluation pipelines, or automated quality assessment.
refactor:scikit-learn
SnakeO
Refactor Scikit-learn and machine learning code to improve maintainability, reproducibility, and adherence to best practices. This skill transforms working ML code into production-ready pipelines that prevent data leakage and ensure reproducible results. It addresses preprocessing outside pipelines, missing random_state parameters, improper cross-validation, and custom transformers not following sklearn API conventions. Implements proper Pipeline and ColumnTransformer patterns, systematic hyperparameter tuning, and appropriate evaluation metrics.
debug:pytorch
SnakeO
Debug PyTorch issues systematically. Use when encountering tensor errors, CUDA out of memory errors, gradient problems like NaN loss or exploding gradients, shape mismatches between layers, device conflicts between CPU and GPU, autograd graph issues, DataLoader problems, dtype mismatches, or training instabilities in deep learning workflows.
refactor:pytorch
SnakeO
Refactor PyTorch code to improve maintainability, readability, and adherence to best practices. Identifies and fixes DRY violations, long functions, deep nesting, SRP violations, and opportunities for modular components. Applies PyTorch 2.x patterns including torch.compile optimization, Automatic Mixed Precision (AMP), optimized DataLoader configuration, modular nn.Module design, gradient checkpointing, CUDA memory management, PyTorch Lightning integration, custom Dataset classes, model factory patterns, weight initialization, and reproducibility patterns.
video-generation
Superconductor
Generate videos using Google Veo models. Create 4-8 second videos from text prompts, with optional image-to-video and video extension capabilities. Veo 3.1 supports native audio generation including dialogue and sound effects.
capa-officer
QuestNova502
Senior CAPA Officer specialist for managing Corrective and Preventive Actions within Quality Management Systems. Provides CAPA process management, root cause analysis, effectiveness verification, and continuous improvement coordination. Use for CAPA investigations, corrective action planning, preventive action implementation, and CAPA system optimization.
discord-admin-py
JungHoonGhae
Discord server administration via inference.sh - Multi-function app for channel, role, member management, messages, and more. Use for Discord bot operations, server management, channel creation, role assignment, and message handling.
image-generation
Superconductor
Generate high-quality images using Google Gemini's Nano Banana Pro image model. Use this skill when you need to create images from text descriptions or transform existing images/videos into new artwork.
text-to-speech
Superconductor
Convert text to natural-sounding speech using Google Gemini TTS models. Supports 30 different voices and 24 languages. Use this skill when you need to generate audio narration, voiceovers, or spoken content from text.
vllm-deployment
stakpak
Deploy vLLM for high-performance LLM inference. Covers Docker CPU/GPU deployments and cloud VM provisioning with OpenAI-compatible API endpoints.
gemini-consultation
Superconductor
Ask questions to Google Gemini AI models. Use this skill when you need a second opinion from another frontier AI model, want to analyze documents (PDFs, images), or need Gemini's perspective on any topic.
architecture-serverless
KentoShimizu
"Serverless architecture workflow for event-driven and bursty workloads using managed compute and platform services. Use when elasticity and reduced platform operations justify managed-service constraints; do not use when workload shape requires long-lived stateful control loops."
Comunicador Marketing
scaleto
Enlace A2A del Grupo Marketing.
design-qa-implementation-parity
KentoShimizu
"Verify implementation parity against approved design specs with severity-based decisions and fix guidance. Use when implemented UI must be compared against approved specs before release or sign-off; do not use for backend data-model or deployment pipeline decisions."
django-app-development
KentoShimizu
"Django application development workflow for production-grade app changes. Use when Django-specific code (models, views, serializers/forms, routing, settings, middleware, auth flows) must be implemented or revised with framework runtime behavior in scope; do not use for repository-wide architecture governance or release management policy."
llm-interview-coach
llx9826
Prepare for large language model, generative AI, and LLM algorithm interviews. Use when the user wants mock interviews, study plans, question drills, answer review, or targeted prep for 大模型算法岗, LLM engineer, GenAI engineer, 算法工程师, RAG, agent, fine-tuning, inference optimization, evaluation, or model training interviews. Triggers include: "大模型面试", "算法面试", "LLM interview", "mock interview", "帮我准备大模型算法岗", "面试官会怎么问", "RAG面试", "微调面试", "推理优化面试".