conda-doctor
Debug Conda environments
TLDR
View report for the currently active environment
Specify an environment by name
Specify an environment by its path
Enable verbose output (Note: the -v flag can be repeated to increase verbosity)
SYNOPSIS
conda-doctor [ENV_PATH] [--debug] [--section SECTION] [--json] [--fix] [-h --help]
PARAMETERS
ENV_PATH
Path to specific Conda environment (default: checks all)
--debug
Enable verbose debug logging
--section SECTION
Limit checks to named section (e.g., prefix, activation, solver)
--json
Emit machine-readable JSON output instead of human-readable report
--fix
Attempt automatic fixes for detected issues (experimental; use cautiously)
-h, --help
Display usage summary and exit
DESCRIPTION
conda-doctor is a diagnostic utility for identifying and troubleshooting problems in Conda installations and environments. It scans environment prefixes for issues like missing files, broken symlinks, incorrect permissions, and activation script failures. The tool also checks solver satisfiability, package integrity, and shell hook configurations, categorizing findings into actionable sections such as prefix, activation, and solver. Ideal for data scientists and developers maintaining complex multi-environment setups, it provides verbose reports to pinpoint root causes of errors like failed activations or dependency conflicts.
Run without arguments to check all environments, or specify a path for targeted analysis. JSON output enables scripting and automation. While primarily for maintenance, its --fix mode offers semi-automated remediation for common issues, enhancing reproducibility in scientific computing workflows.
CAVEATS
Third-party tool, not official Conda component.
--fix modifies environments; always backup first.
May require elevated privileges for system-wide checks.
INSTALLATION
pip install conda-doctor or conda install -c conda-forge conda-doctor
COMMON SECTIONS
prefix: File/symlink integrity.
activation: Shell hook validation.
solver: Dependency resolution tests.
HISTORY
Developed as open-source project in 2021 by Conda community contributors. Hosted on conda-forge/PyPI; evolved to support Conda 4.10+ and multi-platform diagnostics.


