# Contributing to photographi
Welcome! `photographi` is a **Local Computer Vision Engine**. We are building the future of privacy-first, programmable photo libraries.
## 🌟 Vision
We believe that:
1. **Privacy is non-negotiable:** Your photos should never leave your machine to be "understood."
2. **Photos are Data:** Every image has latent signal (sharpness, noise, subject) that should be queryable.
3. **Workflows are Personal:** We provide the engine; users build the workflows (Search, Culling, Auditing).
## 🛠️ Development Setup
### 1. The Stack
* **Core Logic**: `photo-quality-analyzer` (Physics + Neural SDK)
* **Application**: `photographi` (CLI + MCP Server)
* **Models**: YOLO26n (Nano Indexing) / YOLO11x (Studio Grade)
### 2. Quickstart (Local Development)
```bash
# 1. Clone the repository
git clone https://github.com/prasadabhishek/photographi.git
cd photographi
# 2. Setup Virtual Environment
python3 -m venv venv
source venv/bin/activate
# 3. Install in Editable Mode
pip install -e .[test]
# 4. Activate Automation Hooks
bash scripts/setup_hooks.sh
```
---
## 🧪 Testing & Verification
We maintain a zero-regression policy. Before pushing any code, you **must** run the local verification suite:
```bash
./verify.sh
```
This runs the full `pytest` suite and ensures cross-platform compatibility. Our **pre-push hook** will automatically block pushes if this suite fails.
---
## 🚀 Release Process
Releases are automated via GitHub Actions:
1. Add release notes to [CHANGELOG.md](CHANGELOG.md) under the `[Unreleased]` or next version header.
2. Run `./scripts/release.sh <version>` to bump version strings and create a git tag.
3. Push the tag: `git push origin v<version>` to trigger the PyPI/GitHub Release pipeline.
---
## 🗺️ Roadmap
We are looking for contributions in:
* **Video Analysis**: Extending the engine to support `.mp4` and `.mov`.
* **Facial Recognition**: Local, privacy-first clustering of faces.
* **New Workflows**: Integrations with Capture One, Lightroom, or Darktable.
---
**License**: MIT
