research template

odf init my-project --template research

default (Item/Items/SQLite) plus a single RunExperiment @Task and a write-up stub, following "Good Enough Practices in Scientific Computing" (Wilson et al., 2017): a flat data/ for raw, read-only input, a results/ for everything generated, and a doc/ for the write-up — app/ plays the role of that layout's src/.

Layout

my-project/
├── config.toml
├── app/
│   ├── __init__.py
│   ├── entities.py       # Item
│   ├── storages.py       # SQLite
│   ├── repositories.py   # Items(SQLite)
│   └── experiments.py    # RunExperiment
├── data/
│   └── items.csv          # source data — read, never edited in place
├── results/                # starts empty — RunExperiment writes here
└── doc/
    └── notes.md            # hypothesis / method / results stub

What's added

# app/experiments.py
@Task
class RunExperiment:
    def execute(self) -> dict:
        if not self.items.all():
            ...  # seeds data/items.csv into Items, first run only

        all_items = self.items.all()
        summary = {
            "item_count": len(all_items),
            "total_quantity": sum(item.quantity for item in all_items),
        }
        ...  # written to results/summary.json
        return summary

doc/notes.md is a lab-notebook stub — hypothesis, method, results, notes — pre-filled with a pointer back to RunExperiment so the write-up and the code that produces its numbers stay next to each other.

Run it

odf runUI running at http://127.0.0.1:4747

Trigger RunExperiment from the UI (or execute_task over MCP), then check results/summary.json and fill in doc/notes.md.