LLM

Pi

stars 104.8Kforks 13.2K

The minimal, self-extensible coding agent for the terminal

magic-compact

stars 167forks 13

Lossless context compression for Claude Code and OpenCode

leanctx

stars 327forks 5

Drop-in LLM token compression. Code stays verbatim; prose compresses.

token-optimizer-mcp

stars 522forks 62

Token savings and persistent knowledge graphs for AI coding agents

OmniRoute

stars 65.3Kforks 9.1K

352-provider AI gateway with automatic fallback and token compression

i-have-adhd

stars 41.9Kforks 2.4K

ADHD-friendly output rules for Claude Code. Action first, no filler.

context-mode

stars 22.7Kforks 1.6K

Cuts LLM context use by up to 98% per session via local MCP sandbox

miles

stars 2.8Kforks 489

Enterprise-grade RL framework for LLM post-training at scale

llama-swap

stars 5.7Kforks 464

Hot-swap AI models on your local inference server

Automated-AI-Web-Researcher-Ollama

stars 3Kforks 276

Ollama-powered web researcher that breaks queries into focus areas

sie

stars 3.3Kforks 312

One self-hosted cluster for all the models your agents need