LLM

Теги: llm 

Drivers

• [notes/llama-cpp]
• web-llm. High-performance In-browser LLM Inference Engine

Specs

• OpenSpec. Spec-driven development (SDD) for AI coding assistants.
• Agent Skills. A standardized way to give AI agents new capabilities and expertise.

Models

• smollm2 SmolLM2 is a family of compact language models available in three size: 135M, 360M, and 1.7B parameters. On hugging face
• GLM (General Language Model)

Models hosts

• ollama Access the latest open models with complete privacy locally or in the cloud. Your prompts are never stored or trained on.

Apps and libs

• spec-kit Toolkit to help you get started with SDD or any other process!
• spek. A lightweight, read-only viewer for OpenSpec content for [lists/vscode]
• openwiki. OpenWiki is a CLI that writes and maintains agent documentation for your codebase.

AI-ide

• zed. Zed is an open-source code editor with built-in collaboration and AI tools. github, zed gdscript extention

Agents systems

• oh-my-openagent OmO: Just type “mass ulw” keyword with your prompt. Now you are the master of graph engineering. github repo
• Paseo - llm оркестратор. Paseo runs your coding agents on your machine and gives you a mobile, desktop, web, and CLI client to drive them from anywhere.
• OpenClaw. Многоагентная экосистема на базе llm. Organizes your inbox, sends emails, manages your calendar, checks you in for flights. All from WhatsApp, Telegram, or any chat app you already use. Run it for yourself — or share it with your team.
• DeepCode. Advancing Code Generation with Multi-Agent Systems
• OpenCode. OpenCode is an open source AI coding agent. It’s available as a terminal-based interface, desktop app, or web app.
• little-coder, A coding agent tuned for small local models, built on top of pi.
• hermes-agent. The self-improving AI agent built by Nous Research. It’s the only agent with a built-in learning loop — it creates skills from experience, improves them during use, nudges itself to persist knowledge, searches its own past conversations, and builds a deepening model of who you are across sessions. Run it on a $5 VPS, a GPU cluster, or serverless infrastructure that costs nearly nothing when idle.