Versioned backups
Export environments to timestamped YAML files, with history-only mode for portable migrations.
Python CLI · Conda · Cross-platform
conda-helper wraps frequent environment tasks into readable one-liners, adds safer defaults, and turns common Conda failures into actionable hints for research, data science, and local development workflows.
$ conda-helper ls
NAME SIZE PREFIX
base 4.2 GB /opt/miniconda3
vision-lab 9.8 GB /envs/vision-lab
$ conda-helper ls --no-size
NAME SIZE PREFIX
base - /opt/miniconda3
vision-lab - /envs/vision-lab
$ conda-helper backup vision-lab --from-history
backup written: vision-lab-20260601-101530-history.yml
$ conda-helper pack vision-lab -o ./dist
archive ready: dist/vision-lab-20260601-101630.tar.gz
Feature map
Designed for users who manage many local environments and want a safer, more memorable interface on top of native Conda.
Export environments to timestamped YAML files, with history-only mode for portable migrations.
Rebuild from any backup and override the target environment name during restore.
Wrap native conda create --clone into a compact, easy-to-remember command.
Prefer conda-pack for relocatable archives and fall back to metadata-rich tarballs.
Batch-remove stale environments and preview cache cleanup with dry-run support.
Run health checks and translate network, disk, channel, and permission problems into practical hints.
Workflow design
conda-helper organizes Conda operations around a repeatable lifecycle: inspect, preserve, move, deploy, and clean. Each command has clear output and conservative behavior by default.
conda-helper ls shows size and prefix; --no-size skips scanning for fast checks.backup --from-history keeps migrations portable.restore, clone, and pack cover common transfer paths.doctor and friendly errors guide next actions.Technical design
The command layer provides ergonomic flags, aliases, validation, and terminal-friendly output.
High-level operations return plain Python data so core behavior can be unit-tested independently.
A focused adapter delegates to the real Conda executable instead of reimplementing package logic.
Known Conda failure patterns are mapped to friendly explanations and next-step suggestions.
Install
Requires Python 3.8+ and an available Conda installation.
pip install conda-helper
conda-helper doctor
# or install from source
git clone https://github.com/wellkilo/conda-helper.git
cd conda-helper
pip install -e .