ivx 0.1.0

Standalone live progress dashboard for AI-led R&D workflows

ivx Dashboard

Chinese documentation: see README_ZH.md.

Theme

Human-in-the-loop operational visibility for AI-driven software delivery.

This project focuses on one core question: How can teams trust and control AI execution flow in real time without losing delivery speed?

Introduction

ivx Dashboard is a standalone runtime board for AI-led R&D workflows. It aggregates project progress, CI/testing health, collaborator status, intervention signals, and recent activity into one live view.

It is designed for practical operations, not just reporting:

• monitor execution state continuously
• detect risk and stale activity early
• trigger human intervention when thresholds are breached
• keep project switching and multi-project tracking consistent

The dashboard includes privacy-oriented defaults such as project path masking to avoid leaking local workstation directories during demos or screenshots.

Project Overview

• Runtime server: server.py
• App entry: app.py
• Web UI: web/index.html
• Local data: data/live_progress.json, data/dashboard_state.json

Installation

1. Before PyPI release:
   • pip install "git+https://github.com/rendao/ivx.git@main"
2. After PyPI release:
   • pip install ivx
3. Pinned production install:
   • pip install "ivx==<version>"

Quick Start

1. Start service:
   • python app.py --host 127.0.0.1 --port 8789
2. Open dashboard:
   • http://127.0.0.1:8789
3. Configure or switch project in the UI:
   • Click the Project button in the header.
   • Fill project name/path, then Save / Switch.
   • Optionally enable Force re-bootstrap when initializing a new project contract.

Privacy and Path Masking

• API/UI output does not expose absolute local directories.
• Project path fields are masked (configured vs hidden), to avoid leaking workstation paths in screenshots or demos.

GitHub and Package Naming Agreement

Use one name everywhere:

• GitHub repository: ivx
• Python package/module: ivx