Starred repositories
GPU kernel optimization for LLM inference (AMD MI355X)
WAVE (Wide Architecture Virtual Encoding) - The universal GPU ISA. Write GPU kernels once, run on Apple, NVIDIA, AMD, and Intel GPUs. Includes compiler, four backends, emulator, and SDKs for Python…
Frequency-based KV cache pruning for llama.cpp — 25% cache reduction, improved PPL at long context. GPU compaction kernel for HIP/ROCm.
this repository contains examples of GPU kernels written in AMD's HIP
Agents, and RL environment, for optimizing GPU kernels on AMD ROCm using LLM agents. Benchmarks LLM serving workloads end-to-end, profiles bottleneck kernels, optimizes them via Claude Code or Code…
🔥 LLM-powered GPU kernel synthesis: Train models to convert PyTorch ops into optimized Triton kernels via SFT+RL. Multi-turn compilation feedback, cross-platform NVIDIA/AMD, Kernelbook + KernelBench
Optimized FP16/BF16 x FP4 GPU kernels for AMD GPUs
A GPU benchmark tool for evaluating GPUs and CPUs on mixed operational intensity kernels (CUDA, OpenCL, HIP, SYCL, OpenMP)
Domain-specific language designed to streamline the development of high-performance GPU/CPU/Accelerators kernels
MCP server for Email (IMAP/SMTP) - composite tools optimized for AI agents
Lucebox optimization hub: hand-tuned LLM inference, built for specific consumer hardware.
This is a Model Context Protocol (MCP) server that gets answers from your Perfetto Traces. It turns natural‑language prompts into focused Perfetto analyses.
Local LLM-assisted text completion for Qt Creator.
Port of OpenAI's Whisper model in C/C++
Tensors and Dynamic neural networks in Python with strong GPU acceleration
Custom ROCm 7.11 stack for gfx103x (independent repo)
This is MCP server for Claude that gives it terminal control, file system search and diff file editing capabilities
An MCP server plus a CLI tool that indexes local code into a graph database to provide context to AI assistants.
MCP server + Chrome extension that gives AI agents control of your real browser with existing sessions and logins
Give your AI agent a real browser — with a human in the loop. Open-source MCP-native browser agent.
Automate any website. AI compiles it. Runs forever at $0. 200+ skills, 3 runtimes, MCP native.
MCP server with multi-tier fallback chain for fetching web content as clean markdown
Smart Tree: not just a tree, a philosophy. A context-aware, AI-crafted replacement for 20+ tools with MEM8 quantum compression, semantic search, AST-smart editing, and partnership memory. Crafted w…
Intelligent code indexing MCP server. 13 tools, 10 languages, hybrid search, call graphs, O(1) symbol retrieval.
Agent-MCP is a framework for creating multi-agent systems that enables coordinated, efficient AI collaboration through the Model Context Protocol (MCP). The system is designed for developers buildi…
Deterministic orchestrator for 30+ CLI AI coding agents. Git worktree isolation, HMAC audit trail, MCP server mode.
Per-component Internationalisation solution for JS application. Type-Safe. Translate with AI. Edit Visually.