Metadata-Version: 2.4
Name: industrial-rag-agent
Version: 0.1.6
Summary: Local RAG agent for Industrial AI Agents — watches folders, extracts text, OCRs scanned PDFs, indexes in SQLite FTS5
Author-email: Pascal Goemaere <pascal.g@rmplating.com>
License-Expression: LicenseRef-Proprietary
Requires-Python: >=3.11
Description-Content-Type: text/markdown
Requires-Dist: watchdog>=4.0
Requires-Dist: python-docx>=1.1
Requires-Dist: openpyxl>=3.1
Requires-Dist: pymupdf>=1.24
Requires-Dist: uvicorn>=0.30
Requires-Dist: httpx>=0.27
Provides-Extra: ocr-win
Requires-Dist: pytesseract>=0.3; extra == "ocr-win"
Provides-Extra: ocr-linux
Requires-Dist: pytesseract>=0.3; extra == "ocr-linux"

# Industrial RAG Agent

Local document indexer for **Industrial AI Agents** SaaS.

Runs on the client's computer — documents **never leave their network**.

## How it works

```
📁 Document Folder
    ↓ (file watcher)
📄 Text Extraction (PDF, DOCX, XLSX, TXT/CSV)
    ↓ (+ OCR for scanned PDFs)
🗃️ SQLite FTS5 Index (local)
    ↓
🌐 Local API (localhost:3847/search)
    ↓
☁️ Industrial AI Agents (cloud) queries this endpoint
```

## Install

```bash
pip install industrial-rag-agent
```

## Quick Start

```bash
# 1. Configure
industrial-rag-agent configure

# 2. Start (watches folders + runs API)
industrial-rag-agent start
```

## Commands

| Command     | Description                              |
|-------------|------------------------------------------|
| `configure` | Interactive setup (API key, folders, port) |
| `start`     | Start watcher + local API server          |
| `index`     | One-time index of all existing documents  |
| `search`    | Search from the command line              |
| `status`    | Show index statistics                     |

## OCR Support

| OS      | Engine                    | Install Required? |
|---------|---------------------------|-------------------|
| macOS   | Apple Vision (Neural Engine) | No — built in    |
| Windows | Tesseract OCR             | Yes               |
| Linux   | Tesseract OCR             | Yes               |

## API Endpoints

- `GET /search?q=keyword&limit=5` — search indexed documents
- `GET /status` — agent status and index stats
- `GET /health` — health check

## Requirements

- Python 3.11+
- Tesseract OCR (Windows/Linux only)
