Packages

npm-native package manager for reusable Agent Skills.

[](https://opensource.org/licenses/MIT)

Team9 Monorepo - Client (Tauri) + Server (NestJS)

Uvicorn is an ASGI web server for Python that runs ASGI applications and provides HTTP/1.1 and WebSocket support.

[](https://platform.claude.com/docs/en/agents-and-tools/agent-skills/overview)

SkillHub is a self-hosted, open-source agent skill registry that lets organizations publish, version, govern, and distribute reusable “skill packages” within private namespaces. It provides a Web UI and CLI-first workflows plus a backend REST API, with RBAC and audit logging, and supports pluggable storage (filesystem, S3/MinIO).

AgentScope is a Python multi-agent framework for building, orchestrating, and running LLM agents with reusable abstractions (agents, toolkits, memory, planning/workflows) and integrations including MCP tool support and additional protocols (e.g., A2A) plus observability via OpenTelemetry.

Claude Code skill for detecting flaky test patterns that cause intermittent CI failures

[](https://github.com/vicentereig/dspy.rb)

Created by **[Anivar Aravind](https://anivar.net)**

Created by **[Anivar Aravind](https://anivar.net)**

fastmcp-builder is a Claude Code skill and reference project that helps you generate production-ready MCP servers using the FastMCP Python framework. It provides reference guides, runnable examples, and a complete reference implementation (including OAuth integration, testing patterns, and a recommended project structure).

OpenRLHF is an open-source RLHF framework for training and improving language models using reinforcement learning from human feedback. It provides distributed RL training (e.g., PPO, REINFORCE++, GRPO, RLOO) built around Ray orchestration and vLLM-based fast sample generation, with support for multi-turn agent-based execution and integration with HuggingFace/DeepSpeed for large-model training.

Provides an open-source “skill” for AI coding/agent platforms that can generate images and videos using CyberBara-backed models (e.g., Nano Banana, Sora, Seedance, Kling/Ve o), including helper commands for credit quotation/balance checks and task waiting to retrieve final media URLs.

LitServe is a lightweight Python framework (FastAPI-based) for building and serving custom AI inference servers. Users implement a LitAPI with `setup()` and `predict()` (and potentially more advanced logic like batching/streaming/routing), then run it via `LitServer` to expose an HTTP API for inference pipelines, including agents, RAG, and multi-model workflows. It supports self-hosting and deployment via Lightning’s cloud offering.

PencilPlaybook is a Claude Code “skill” intended to guide Claude in producing UI/design outputs inside Pencil.dev by embedding perceptual design defaults (token lookup tables), structured workflows for working with .pen canvases, and presets/scaffolds for common screen archetypes. It also references a setup wizard that configures design-system presets and claims to provide a complete parameter reference for 12 Pencil MCP tools.

Agno is a Python runtime/framework for building, running, and managing agentic software at scale. It provides an agent framework with tools and integrations (including MCP tools), a production runtime using a stateless, session-scoped FastAPI backend, and a control plane/UI (AgentOS) for testing, monitoring, and management.

Language: English | [日本語](README.ja.md)

⚠️ Better agent skill for mac [flomo-local-api-skill](https://github.com/Undertone0809/flomo-local-api-skill)