Setting Up Your Development Environment#

Clone and Fork

git clone https://github.com/<your-username>/data-curator.git
cd data-curator
git remote add upstream https://github.com/data-curator/data-curator.git
git fetch upstream

Install PDM and Development Dependencies

pip install pdm
pdm install
pdm run install_dev

(Optional) Create a Virtual Environment Although PDM manages environments automatically, if you prefer a venv or Conda, activate it before running pdm install.
Install Data Curator in Editable Mode
```
pdm run install_dev
```
Install the AI scaffolding for your AI coding assistant You can check the available targets at microsoft/apm The following example is for Claude:
```
apm config set target claude
apm install
```
Verify the Test Suite
```
pdm run test
```
Run the Linter
```
pdm run lint
```

Branching and Pull Request Workflow#

Users without write access must fork the repository; those with write access may create branches directly, always based on the dev branch.

Forking and Synchronizing#

Always keep your local dev branch in sync:

git checkout dev
git pull upstream dev --ff-only

Creating an Issue Branch#

When contributing, all contributions must be based on an existing registered issue on the Data Curator GitHub repository. When created, each issue gets assigned an issue number, and the contribution branch name should be of the form issues/<issue-number> where <issue-number> is the number of the issue.

```
git checkout -b issues/<issue-number>
```
Example:

git checkout -b issues/4