New Modular Project

modular/skills/new-modular-project

作者 modularb9b3a8e86700無授權條款200 個星標收錄於 2026年10月8日更新於 2026年10月8日儲存庫今天更新

Creates a new Mojo or MAX project. Use when wanting to start a new Mojo or MAX project, initializing the `pixi` or `uv` environment to use Mojo or MAX, or when the user wants to begin a new Mojo or MAX project from scratch.

AI 產生的概覽

指導建立新的 Mojo 或 MAX 專案,選擇環境管理器與發行通道。

功能
此技能引導初始化新的 Mojo 或 MAX 專案。它會蒐集未指定的選項,例如專案名稱、專案類型、環境管理器與發行通道,接著提供 pixi、uv、pip 或 conda 的設定指令。它也涵蓋系統前置條件,例如 C 連結器,以及 MAX 與 Mojo 之間的版本對齊。
適用情境
適用於從零開始建立新的 Mojo 或 MAX 專案,或為 Mojo、MAX 設定 pixi、uv 環境時。也適用於檢查 MAX 與 Mojo 是否來自同一個通道。
執行需求
僅為說明內容,技能未附帶指令碼。需要命令列環境,可能需要安裝 pixi、uv、pip 或 conda,以及 gcc 或 Xcode 命令列工具等 C 連結器。存取套件索引與文件需要網路連線。

When the user wants to create a new project, first infer as many options as possible from the user's request (for example, "new Mojo project" means type=Mojo, "called foo" means name=foo). Then use a structured multiple-choice prompt (not plain text) to gather only the remaining unspecified options in a single interaction. Do NOT ask about options the user has already provided or implied. The options to determine are:

  • Project name: ask if the user hasn't specified one.
  • Type of project: Mojo or MAX (infer from context if the user said "Mojo project" or "MAX project").
  • Environment manager: pixi (recommended) or uv.
  • uv project type (only when the environment manager is uv): full uv project (uv init + uv add, recommended) or quick uv environment (uv venv + uv pip install, lighter weight).
  • Channel: nightly or stable. Default to nightly for MAX projects and stable for Mojo projects, and ask only if the user hasn't implied one.

Then follow the appropriate section below (pixi or uv) to initialize the project and choose max or mojo as appropriate. Don't pin a version: each channel already resolves to the right one.

MAX and Mojo ship together but number their releases differently, so their version strings don't look alike. On the stable channel, max is 26.5 while mojo is 1.0.0; on nightly they're 26.6.0.dev* and 1.1.0.dev*. That's expected, not a mismatch.

[!NOTE] Mojo and MAX projects use pixi. The magic CLI is retired; its replacement is the modular CLI. Don't use or suggest magic.


System prerequisites

Mojo requires a C linker for compilation. Install one if not already present:

OSCommand
Ubuntu/Debiansudo apt install gcc
Fedora/RHELsudo dnf install gcc
macOSxcode-select --install
WindowsInstall WSL2 first (see Windows users), then install gcc

Windows users: Mojo doesn't run natively on Windows. Install WSL2 (wsl --install in PowerShell), then follow the Linux instructions inside your WSL environment.


Pixi (recommended)

Pixi manages Python, Mojo, and other dependencies in a reproducible manner inside a controlled environment.

First, determine whether pixi is installed. If it isn't available at the command line, install it using the latest instructions at https://pixi.prefix.dev/latest/#installation.

After installing pixi, you may need to add it to the local shell environment.

Nightly

bash
# New projectpixi init [PROJECT] \  -c https://conda.modular.com/max-nightly/ -c conda-forge \  && cd [PROJECT]pixi add [max / mojo]pixi shell
# Existing project - add to pixi.toml channels first:# [workspace]# channels = ["https://conda.modular.com/max-nightly/", "conda-forge"]pixi add [max / mojo]

Stable

bash
# New projectpixi init [PROJECT] \  -c https://conda.modular.com/max/ -c conda-forge \  && cd [PROJECT]pixi add [max / mojo]pixi shell
# Existing projectpixi add [max / mojo]

Python-using projects

If your project uses Python libraries with Mojo:

bash
pixi add pythonpixi add requests           # conda-forge packagespixi add --pypi some-pkg    # PyPI-only packages

uv

uv is a fast and very popular package manager, familiar to developers coming from a Python background. It also works well with Mojo projects.

Nightly (project)

bash
uv init [PROJECT] && cd [PROJECT]uv add [max / mojo] \  --index https://whl.modular.com/nightly/simple/ \  --prerelease allow

Stable (project)

bash
uv init [PROJECT] && cd [PROJECT]uv add "max[all]"

This resolves from PyPI. The all extra pulls in mojo and the rest of the MAX stack, so the MAX and Mojo versions always match.

Nightly (quick environment)

bash
mkdir [PROJECT] && cd [PROJECT]uv venvuv pip install [max / mojo] \  --index https://whl.modular.com/nightly/simple/ \  --prerelease allow

Stable (quick environment)

bash
mkdir [PROJECT] && cd [PROJECT]uv venvuv pip install "max[all]"

When using uv, you can use max or mojo directly by working within the project environment:

bash
 source .venv/bin/activate

pip

Standard Python package manager.

Nightly

bash
python3 -m venv .venv && source .venv/bin/activatepip install --pre [max / mojo] \  --extra-index-url https://whl.modular.com/nightly/simple/

Use --extra-index-url, not --index-url. The latter replaces PyPI, and the nightly index doesn't carry third-party dependencies like numpy, so pip backtracks through every max version instead of reporting a clear error.

Stable

bash
python3 -m venv .venv && source .venv/bin/activatepip install "max[all]"

As with uv, the all extra installs mojo as well, so the MAX and Mojo versions always match.


Conda

For conda and mamba users.

Nightly

bash
conda install -c conda-forge \  -c https://conda.modular.com/max-nightly/ [max / mojo]

Stable

bash
conda install -c conda-forge \  -c https://conda.modular.com/max/ [max / mojo]

Version alignment with MAX

If using MAX with custom Mojo kernels, both must come from the same channel. Don't compare their version numbers: MAX and Mojo number releases differently, so a matching pair looks mismatched (stable is max 26.5 with mojo 1.0.0).

bash
# Check that both came from the same channelpixi list | grep -E "^(max|mojo)\b"

Or, instead install max[all] (with pip/uv) or max-all (with conda/pixi):

bash
uv add "max[all]"
bash
pixi add max-all

Installing max with "all" optional dependencies instead of installing max and mojo separately will ensure that the max and mojo versions always match. Mixing versions between the two causes kernel compilation failures.


References

來源與署名

來源:modular/skills位於new-modular-project提交b9b3a8e

授權條款: 無授權條款

內容歸原作者所有。SourceWeft 從公開儲存庫中收錄這些內容。

檢舉或申請下架