cli-workflow.md

CLI Workflow

The package exposes the spt command, defined in src/simulation_project_template/cli.py and registered in pyproject.toml:

[project.scripts]
spt = "simulation_project_template.cli:main"

Built-in commands

# Generate N random numbers in [low, high) and save to JSON.
spt generate <count> [-o OUTPUT] [--low 0.0] [--high 1.0] [--seed SEED]

# Compute the mean of a previously generated file.
spt mean <input.json>

Examples

spt generate 1000 -o data/sample.json --seed 42
spt mean data/sample.json

Adding a new command

Commands are plain functions registered on an argparse subparser. No global registry or plugin system is needed — add a function, wire it up, and it appears in spt --help.

# src/simulation_project_template/cli.py

def _cmd_mycommand(args: argparse.Namespace) -> int:
    # implement your command here
    print(args.value)
    return 0


# Inside main(), after the other sub.add_parser(...) blocks:
cmd = sub.add_parser("mycommand", help="Short description.")
cmd.add_argument("value", help="Some input.")
cmd.set_defaults(func=_cmd_mycommand)

The command is then available as:

spt mycommand <value>

Snakemake scripts vs. the CLI

Snakemake scripts in workflow/scripts/ import directly from the package (from simulation_project_template import ...) — they do not go through the CLI. The CLI is intended for local interactive use and quick sanity checks, not for pipeline execution.