Agent Skills Hub

pytorch-lightning

PyTorch Lightning skill for scalable deep learning: automates model training, multi-GPU orchestration, data pipelines, and distributed training strategies like DDP, FSDP, and DeepSpeed.

flow-nexus-neural

Train and manage neural networks in distributed E2B sandboxes using the Flow Nexus platform, supporting custom architectures like Transformers, LSTMs, and GANs.

speak

Local text-to-speech conversion using Kokoro TTS. Generate audio, read text aloud, and handle multilingual speech synthesis directly in your terminal.

stable-baselines3

Production-ready reinforcement learning using Stable Baselines3. Train agents, design custom environments, implement training callbacks, and optimize workflows with a scikit-learn-style API.

debug-distributed

Debugging guide for AReaL distributed training issues, including hangs, NCCL errors, OOM, and numerical consistency in FSDP2/TP/CP/EP.

llmfit-advisor

Analyze local system hardware (RAM, CPU, GPU/VRAM) to receive expert recommendations for optimized local LLM models, quantization settings, and performance estimates.

pyhealth

Comprehensive Python healthcare AI toolkit for clinical data processing, medical coding translation, and developing deep learning models like RETAIN and Transformers for EHR, physiological signals, and clinical prediction tasks.

mls

Unified local ML inference server for ASR, TTS, Translation, Image Generation, and Vision on Apple Silicon, powered by MLX.

generate-image

Generate or edit images using AI models like FLUX and Gemini. Ideal for photos, illustrations, concept art, and visual assets, excluding technical diagrams and schematics.

debug-cuda-crash

Tutorial for identifying and resolving CUDA runtime crashes using FlashInfer's API logging framework.

transcription

Production-ready audio/video transcription using OpenAI Whisper. Features model selection, timing synchronization, speaker diarization, and batch processing for media workflows.

gemini-video-understanding

Perform advanced video analysis using Google's Gemini API: summarize content, transcribe audio, extract timestamps, clip segments, and analyze YouTube URLs or local files with support for multiple models and long contexts.