project_init

Scaffold a new EvolvingSystems research project on disk.

Emits the locked directory layout (see docs/DESIGN.md - researcher-project layout) so every project the training-decider agent bootstraps has the same predictable shape:

<name>/ pyproject.toml README.md .gitignore data/ README.md raw/.gitkeep fetch/.gitkeep process/.gitkeep datasets/.gitkeep benchmark/.gitkeep src/ init.py verifiers.py metrics.py transforms.py experiments/.gitkeep

Use via the CLI: evsys init-project \<name>. Programmatically: call init_project(path, name=...).

attributeSCAFFOLD_DIRS

= ('data', 'data/raw', 'data/fetch', 'data/process', 'data/datasets', 'data/benchmark', 'data/validation', 'src', 'experiments')

attributeGITKEEP_DIRS

= ('data/raw', 'data/fetch', 'data/process', 'data/datasets', 'data/benchmark', 'data/validation', 'experiments')

attribute__all__

= ['GITKEEP_DIRS', 'SCAFFOLD_DIRS', 'init_project']

funcinit_project(target, *, name=None, force=False) -> Path

Create a new research project skeleton under target.

If target exists and is non-empty, refuse unless force=True. Returns the absolute project path.

paramtargetstr | Path

paramnamestr | None

= None

paramforcebool

= False

Returns

pathlib.Path

func_write(path, content) -> None

parampathPath

paramcontentstr

Returns

None

func_pyproject_toml(name) -> str

paramnamestr

Returns

str

func_readme(name) -> str

paramnamestr

Returns

str

func_gitignore() -> str

Returns

str

func_data_readme() -> str

Returns

str

func_src_init(name) -> str

paramnamestr

Returns

str

func_src_verifiers() -> str

Returns

str

func_src_metrics() -> str

Returns

str

func_src_transforms() -> str

Returns

str