Skip to content

v1.0.0 — Initial Release

Latest
Latest

Choose a tag to compare

View all tags
@tsunamayo7 tsunamayo7 released this 27 Mar 17:35
· 13 commits to main since this release
v1.0.0
cffa4ee

helix-pilot v1.0.0

GUI automation MCP server powered by local Vision LLM (Ollama).

Features

  • 15 MCP tools for comprehensive GUI automation (screenshot, click, type, hotkey, scroll, describe, find, verify, auto, browse, etc.)
  • 100% local — runs via Ollama with zero cloud API cost
  • Windows-native — direct host OS control via Win32 API
  • MCP-compatible — works with Claude Code, Codex CLI, Cursor, VS Code, Open WebUI
  • Safety system — action policies, secret detection, emergency stop, user activity detection

Supported Vision Models

  • mistral-small3.2, gemma3:27b, llava, moondream, or any Ollama vision model

Getting Started

ollama pull mistral-small3.2
git clone https://github.com/tsunamayo7/helix-pilot.git
cd helix-pilot && uv sync

See README for MCP client configuration.

Assets 2
Loading