Agent Skills Hub

ring:testing-skills-with-subagents

A meta-skill for building robust AI agent skills using a TDD approach: define failure (RED), implement the skill (GREEN), and plug rationalization loopholes (REFACTOR).

make-skill-template

Scaffold and generate new GitHub Copilot Agent Skills. Provides templates, directory structures, and instructions to build specialized AI capabilities with bundled resources.

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.

creating-opencode-agents

Expert guidance and configuration standards for creating specialized OpenCode AI agents, including YAML frontmatter, tool permissions, and operational modes.

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.

tool-design

Expert framework for designing agent-facing tools, optimizing tool descriptions, enforcing contract-based APIs, and implementing architectural reduction for reliable AI agent tool selection.

skill-template

A standardized template for creating and documenting modular Agent Skills to ensure consistent, efficient context engineering across AI agent systems.

context-degradation

Diagnose, isolate, and mitigate LLM context failures like lost-in-middle, poisoning, distraction, and context clash to improve agent reliability.

context-fundamentals

Foundational guidelines for context engineering: optimizing token budgets, attention mechanics, and system architecture for AI agents.