getting-started

Getting Started

English | 中文

This guide gets you from a fresh clone to a first loushang code run.

Requirements

• Python 3.11 or newer.
• A model provider credential for online runs.
• A terminal environment that can run Python virtual environments.

Install From Source

git clone https://github.com/zhnt/loushang.git
cd loushang

uv venv .venv
source .venv/bin/activate
uv pip install -e ".[dev]"

Equivalent Makefile shortcut:

make bootstrap
source .venv/bin/activate

make bootstrap creates .venv/ with uv and installs the project in editable development mode. There is no make install target today; make install-binary is reserved for building and installing a local binary.

Check The CLI

loushang --help
loushang --list-models
loushang --list-commands

Run A First Prompt

loushang -p "Inspect this repository and summarize what it does."

Use --model or provider-specific environment variables when you need to select a concrete model route. --model provider:model is a short form (provider/model is also accepted): it works only when that provider/model pair matches exactly one endpoint. If the same provider/model exists under multiple endpoints, the CLI reports an ambiguity and lists explicit provider:endpoint:model alternatives. Use --model provider:endpoint:model when you need to choose a specific endpoint, region, lane, or protocol. In catalog keys, provider and model ids cannot contain :, while endpoint ids may contain :. Project and example model catalog files can be placed under .loushang/models/ or passed explicitly where supported.

Next Steps

• Read the User Guide for sessions, commands, tools, extensions, methods, and diagnostics.
• Read the Examples page for runnable coding and AI SDK scenarios.
• Read the Reference page when you need exact CLI and configuration surfaces.