Metadata-Version: 2.4
Name: verra-one
Version: 0.3.0
Summary: Chat with your business data — emails, docs, spreadsheets.
Project-URL: Homepage, https://github.com/connorberghoffer/verra-one
Project-URL: Repository, https://github.com/connorberghoffer/verra-one
Project-URL: Issues, https://github.com/connorberghoffer/verra-one/issues
Author: Connor Berghoffer
License-Expression: MIT
License-File: LICENSE
Keywords: agent,ai,business,documents,email,llm,rag,search
Classifier: Development Status :: 4 - Beta
Classifier: Intended Audience :: End Users/Desktop
Classifier: License :: OSI Approved :: MIT License
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.11
Classifier: Programming Language :: Python :: 3.12
Classifier: Programming Language :: Python :: 3.13
Classifier: Topic :: Office/Business
Classifier: Topic :: Text Processing :: Indexing
Classifier: Typing :: Typed
Requires-Python: >=3.11
Requires-Dist: chromadb<2.0,>=1.0
Requires-Dist: click<9.0,>=8.1
Requires-Dist: fabric<4.0,>=3.2
Requires-Dist: fastapi<1.0,>=0.115
Requires-Dist: google-api-python-client<3.0,>=2.160
Requires-Dist: google-auth-oauthlib<2.0,>=1.2
Requires-Dist: litellm<2.0,>=1.60
Requires-Dist: msal<2.0,>=1.28
Requires-Dist: openpyxl<4.0,>=3.1.5
Requires-Dist: prompt-toolkit<4.0,>=3.0
Requires-Dist: pydantic<3.0,>=2.10
Requires-Dist: pypdf<5.0,>=4.0
Requires-Dist: python-docx<2.0,>=1.1
Requires-Dist: pyyaml<7.0,>=6.0
Requires-Dist: rich<14.0,>=13.9
Requires-Dist: tiktoken<1.0,>=0.9
Requires-Dist: uvicorn<1.0,>=0.34
Requires-Dist: watchdog<7.0,>=6.0
Provides-Extra: pymupdf
Requires-Dist: pymupdf<2.0,>=1.25; extra == 'pymupdf'
Description-Content-Type: text/markdown

# Verra One

Chat with all your business data — emails, documents, spreadsheets. Runs locally with Ollama or connects to Claude/OpenAI.

## Install

```bash
curl -fsSL https://raw.githubusercontent.com/ConnorBerghoffer/verra-one/main/install.sh | sh
```

Or manually:

```bash
pipx install verra-one    # recommended
pip install verra-one     # if you don't have pipx
```

Then run `verra` — first run walks you through setup.

## Update

```bash
pipx upgrade verra-one    # if installed with pipx
pip install -U verra-one  # if installed with pip
```

## What it does

- **Ingest** folders, Gmail, Google Drive, Outlook — keeps syncing in the background
- **Search** across everything with hybrid retrieval (vector + metadata + keyword)
- **Chat** with your data — answers cite sources, refuses to guess
- **Briefing** — surfaces stale leads, overdue commitments, expiring contracts
- **Actions** — draft emails, set reminders, take notes through the chat
- **Deploy** — `verra deploy user@server` sets up everything remotely via SSH

## Commands

```
verra                       # chat
verra ingest ~/Documents    # add a folder
verra gmail you@gmail.com   # connect Gmail
verra search "pricing"      # search without LLM
verra briefing              # what needs attention
verra serve --port 8484     # HTTP API for building UIs
verra status                # what's ingested
verra delete --source email # remove data
```

Full list: `verra --help`

## API

`verra serve` starts a FastAPI server with endpoints for chat, search, documents, entities, and briefing. OpenAPI docs at `/docs`.

## Models

Works with any LLM via [LiteLLM](https://github.com/BerriAI/litellm). Recommended local model: `llama3.1:8b` via Ollama.

## PDF extraction

Uses [pypdf](https://github.com/py-pdf/pypdf) (MIT) by default. For better quality on complex PDFs:

```bash
pip install "verra-one[pymupdf]"
```

## License

MIT
