Env#
Examples for _env, the environment builder behind
scikitplot env: describe a Python environment once, then build the same
environment on any machine with whichever tool it has (uv, venv + pip,
pipenv, micromamba, mamba or conda).
request arguments, a scikitplot-env.toml, or both (arguments win)
↓
plan reads only: every command, its network use, its approvals
↓
approvals asked (or refused) before the first step runs
↓
execute the plan's steps, under a lock on the environment
↓
receipt what was built; the next run with the same request is a no-op
Every example runs from the command line and from Python with the same fields; the command line form is shown next to each Python call:
scikitplot env plan -n analysis --package "pandas>=2" --verify-import pandas
scikitplot env create -n analysis --package "pandas>=2" --verify-import pandas
The examples need no network: they plan with any backend, and build only
venv environments from the standard library. Named environments go to
SKPLT_ENV_HOME (default ~/.scikitplot/env); the examples point it at a
temporary folder and remove what they build.
Start with the plan, then build, run and remove, then configuration files, policies and the errors an automation reads. See also Environment builder.
Build, Use and Remove an Environment, Idempotently
