-
GitHub Spotlight: ha-mcp – Control Your Smart Home with AI Context
Connect Claude and other AI models directly to your smart home setup using ha-mcp, an open-source Model Context Protocol server for Home Assistant.
-
GitHub Spotlight: hermes-agent – The AI Agent That Learns and Grows With You
Most AI agents forget who you are the moment a session ends. Hermes Agent from Nous Research changes that with built-in memory, autonomous skill creation, and multi-platform support.
-
GitHub Spotlight: PyTorch – Dynamic Neural Networks & Strong GPU Acceleration
Discover how PyTorch turns complex neural network research into fast, flexible Python code with dynamic computational graphs and native GPU acceleration.
-
GitHub Spotlight: LobeHub – Your 24/7 Chief Agent Operator
LobeHub turns standalone LLM prompts into a fully managed, 24/7 autonomous AI workforce. Learn how to hire, schedule, and orchestrate multi-agent workflows at scale.
-
GitHub Spotlight: ODS – Turn Any PC, Mac, or Linux Machine Into a Local AI Server
ODS (Osmantic Deployment System) packages Ollama, Open WebUI, n8n, ComfyUI, and RAG into a unified, zero-cloud local AI stack on your hardware.
-
GitHub Spotlight: LMCache – Supercharge Your LLM with the Fastest KV Cache Layer
Learn how LMCache solves the LLM long-context bottleneck by offloading and sharing Key-Value (KV) caches across vLLM engine instances to crush Time-To-First-Token (TTFT).
-
GitHub Spotlight: FlashInfer – Turbocharging LLM Serving with High-Performance GPU Kernels
Learn how FlashInfer delivers high-throughput GPU kernels for LLM serving engines, supporting DeepSeek MLA, FP4/FP8 compute, and hardware from Turing to Blackwell.
-
GitHub Spotlight: Langfuse – Open Source LLM Observability & Engineering Platform
Build, trace, evaluate, and optimize production LLM applications with Langfuse, the YC-backed open-source LLM observability platform.
-
GitHub Spotlight: hyperframes – Write HTML. Render Video. Built for Agents.
HyperFrames bridges web design and video production, allowing developers and AI agents to transform standard HTML, CSS, and animations into crisp MP4 videos.
-
GitHub Spotlight: Dify – Build Production-Ready AI Agents & RAG Pipelines
Dify turns chaotic LLM development into a unified visual workspace. Build, test, and deploy AI agents and RAG pipelines without rewriting your tech stack.