NC

nibor1896/crow

Developer tools
38 stars 2 forks Quality 62 Trend 26MIT

CROW - Your AI Agent. MCPs, OpenRouter, Any Model or local. It's your choice.

Overview

A local model at 200k context with 25 tools and MCP, persistent memory, its own skills, a browser panel, eyes, and subagents it can send out while it keeps working. Runs on this machine, or on a provider you choose. MoE 512 experts, 10 active 200k context, one slot 73.45 GiB model on disk 30,984 MiB VRAM in use 41.76 tok/s decode 727.65 tok/s prefill yes vision Decode and prefill: 2026-09-01 (#182), driver 616.56, one 33,494-token cold turn per boot, three rounds interleaved against the previous placement (539.98 / 35.74); decode range 40.34–42.76. VRAM settled after load. Conditions in full, and the Linux line: operating points . Preflight, download, per-file sha256 against the release manifest, then the start lines with paths resolved. No elevation, nothing in Program Files, the registry or PATH. Five steps, a per-file sha256 manifest it re-reads on the next run, idempotent.

README

Linux

curl -fsSL https://raw.githubusercontent.com/nibor1896/Crow/main/install.sh | bash

Image tools, Linux (optional: builds sd-server, ~3 min)

curl -fsSL https://raw.githubusercontent.com/nibor1896/Crow/main/install.sh | bash -s -- --build-image-server

Linux, one window (Crow, the crow-nest engine and the operating points you pick; needs gtk3, webkit2gtk-4.1, python-gobject)

curl -fLO https://github.com/nibor1896/Crow/releases/download/v4.3.0/CrowSetup-linux-x64 && chmod +x CrowSetup-linux-x64 && ./CrowSetup-linux-x64

Windows, one window (Crow, the crow-nest engine and the operating points you pick)

irm https://github.com/nibor1896/Crow/releases/latest/download/CrowSetup.exe -OutFile CrowSetup.exe; .\CrowSetup.exe

Windows, Crow alone (image server included in the package)

irm https://raw.githubusercontent.com/nibor1896/Crow/main/install.ps1 | iex

Claude subscription (optional: Claude Code signs in, Crow never sees a token; then Settings → Model → Claude (subscription), or pick it in the operating-point window, details)

npm install -g @anthropic-ai/claude-code && claude auth login

User guide · Reference · Operating points and measurements · Developer guide · Plans · Archive

MIT · nibor1896/crow · Model: Qwen3.8-Flash-Next CNQ4.5-M (Qwen Community License 1.0) · Model: GLM-5.3-Flash (MIT) · Engine: crow-nest · Images: Qwen-Image 2.1 (Qwen Research License, non-commercial) · GGUF line: Unsloth, llama.cpp

View the full README on GitHub

Install

This server does not publish a one-line install command.

Open the repository installation guide New to MCP? Start with the mcpskill.org guide →