RightNow-AI/AutoMegaKernel

An agent harness that compiles a model into one provably-correct, self-retargeting CUDA megakernel and self-tunes it past cuBLAS at batch-1 LLM decode, paper: https://arxiv.org/abs/2606.09682

137Stars on the repository
7Mods indexed here, across every type
2mo agoLast push, which is what freshness is scored on
MITLicence, which decides whether bodies are shown

amk-autoresearch

01

RightNow-AI/AutoMegaKernel

Command Claude Code

Launch the unattended / overnight AMK autoresearch driver on $ARGUMENTS (model [gpu] [minutes|iters]).

137 +2 2mo ago A 25 tokens original MIT

amk-compile

02

RightNow-AI/AutoMegaKernel

Command Claude Code

One-shot compile + verify a model into a CUDA megakernel via amk compile on $ARGUMENTS (model [gpu]).

137 +2 2mo ago A 29 tokens original MIT

amk-optimize

03

RightNow-AI/AutoMegaKernel

Command Claude Code

Drive an interactive AMK propose -> eval -> keep/revert megakernel schedule session on $ARGUMENTS (model [gpu]).

137 +2 2mo ago A 27 tokens original MIT