smithery.ai

conda

Instructions for correct conda usage in opencode

First seen Mar 21, 2026

Installation

$ npx skills add https://smithery.ai

Similar popular skills

Related neighbors and high-traction skills in the same topics — useful to compare before installing.

Also in this package

Other skills from smithery.ai · top by installs.

npx skills add https://smithery.ai

Browse all from smithery.ai

More details

Agent compatibility

Declared targets from SKILL.md / docs. Unmarked agents are not listed — the skill may still install via the CLI.

Claude Code Not declared
Cursor Not declared
Codex Not declared
GitHub Copilot Not declared
Windsurf Not declared
Gemini CLI Not declared
Cline Not declared
OpenCode Declared

Skill metadata

Parsed from SKILL.md frontmatter.

Declared agents opencode

Package contents

Files included with this skill beyond the listing page.

  • skill md SKILL.md 1,691 B
  • docs SUMMARY.md 61 B

History

  1. First seen on skills.sh
  2. First recorded snapshot · 1 installs

SKILL.md

What I do

Correctly utilize conda environments to run python code

When to use me

When you are asked to run python code. You should ask the user which conda enviroment is the correct one to use, and then correctly use the commands listed below to accomplish the task.

When using conda commands in opencode, always prepend the sourcing of conda's profile script to ensure the shell has conda's functions loaded. Opencode runs bash commands in non-interactive shells, which do not automatically source ~/.bashrc or load conda initialization.

For any conda command, such as activation, installation, or environment management, use this format: source ~/miniconda3/etc/profile.d/conda.sh && conda <command> <args>

Examples:

  • To activate an environment: source ~/miniconda3/etc/profile.d/conda.sh && conda activate arxiv
  • To install a package: source ~/miniconda3/etc/profile.d/conda.sh && conda install numpy
  • To create an environment: source ~/miniconda3/etc/profile.d/conda.sh && conda create -n myenv python=3.9

This ensures conda works correctly despite the non-interactive shell limitation. For ease of ensuring consistency with the code, you should never try to install a package nor create an environment without explicit instructions. You can activate environments, but you should first clarify with the user that the environment is the correct one. You should assume that all necessary packages are installed in the conda environment that the user specifies. NEVER try to alter PYTHONPATH. The project should always be configured to correctly run things. If you have questions, ask the user.