LLM Toolkit
一个 Python 命令行工具,统一调用多个大模型 API,支持流式输出、Token 统计、多 Provider 切换。
✨ 功能特性
- 🚀 多 Provider 统一接口:支持 MiniMax、DeepSeek、Anthropic(Mock),新增 Provider 只需写一个类
- 💬 流式输出:边生成边打印,字一个字出来
- ⏱️ 超时 +错误重试:网络错误自动重试,超时 30 秒
- 📝 日志:自动记录调用模型和耗时
- 🔌 CLI 设计:基于 Click,支持参数和子命令
📁 项目结构
src/llm_toolkit/
├── __init__.py
├── cli/
│ ├── __init__.py
│ └── main.py # CLI 入口
├── providers/ # 多 Provider 支持
│ ├── __init__.py # 注册表
│ ├── base.py # 抽象基类
│ ├── minimax.py # MiniMax Provider
│ ├── deepseek.py # DeepSeek Provider
│ └── anthropic.py # Anthropic Provider(Mock)
├── client.py # 预留:统一 LLMClient(Day 6+)
├── models.py # 预留:请求/响应模型
└── config.py # 预留:配置加载
🚀 安装
1. 克隆仓库
git clone <repo-url>
cd m1
2. 安装依赖
需要先安装 uv。
uv sync
⚙️ 配置
在项目根目录创建 .env 文件:
# MiniMax(国内节点)
MINIMAX_API_KEY=sk-cp-xxx
# DeepSeek
DEEPSEEK_API_KEY=sk-xxx# Anthropic(可选,目前未启用)
# ANTHROPIC_API_KEY=sk-ant-xxx
📖 使用
默认 Provider(MiniMax)
llm chat "你好"
切换 Provider
llm chat --provider deepseek "你好"
流式输出
llm chat --stream "写一首关于 Python 的诗"
通过环境变量设置默认 Provider
# PowerShell
$env:LLM_PROVIDER="deepseek"
llm chat "你好"
# Bash
LLM_PROVIDER=deepseek llm chat "你好"
查看帮助
llm --help
llm chat --help
🧪 开发
运行测试
uv run pytest
添加新 Provider
- 在
src/agent/providers/下新建<provider>.py - 继承
BaseProvider,实现create_client() - 在
providers/__init__.py注册
示例:
from .base import BaseProvider
from openai import OpenAI
class MyProvider(BaseProvider):
name = "myprovider"
model = "my-model"
def create_client(self) -> OpenAI:
api_key = os.getenv("MY_API_KEY")
return OpenAI(api_key=api_key, base_url="https://api.my.com/v1")
⚠️ 已知限制
- Anthropic Provider 暂未实现(无 API Key,需要时再填真实代码)
- OpenAI Provider 未集成(无 API Key,需要时按上面示例添加)
- 当前只有
chat命令,后续会加conversation(历史)、prompt(模板)
📝许可证
MIT
Metadata
Release files for chujie-llm-toolkit 0.1.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| chujie_llm_toolkit-0.1.0.tar.gz | 11.6 kB | Without provenance | Details for chujie_llm_toolkit-0.1.0.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| chujie_llm_toolkit-0.1.0-py3-none-any.whl | Python 3 | none | any | Details for chujie_llm_toolkit-0.1.0-py3-none-any.whl |
Total release size: 23.2 kB
Release files / chujie_llm_toolkit-0.1.0.tar.gz
| Download URL | chujie_llm_toolkit-0.1.0.tar.gz |
|---|---|
| Size | 11.6 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
3a0843905bbdbe7dd1f8aa0e260725aaca092c9031c96e0af8348a47e5d1dcc7
|
|
BLAKE2b-256 checksum How to use checksums |
004a56aab9060f9d4e21d7f905c2dfeb3570b8ee01dda687088f78dff9018cc2
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
uv/0.12.12 {"installer":{"name":"uv","version":"0.12.12","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":null,"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":null}
|
Release files / chujie_llm_toolkit-0.1.0-py3-none-any.whl
| Download URL | chujie_llm_toolkit-0.1.0-py3-none-any.whl |
|---|---|
| Size | 11.5 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
3a820b4771d899aa1111117adc1ca271a9017490f553a415844dbc71ca30636b
|
|
BLAKE2b-256 checksum How to use checksums |
79c219570a6f1c0d77ace87675c60f2c97d9fee5ccfb08f7f6d90e5baf6e72b1
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
uv/0.12.12 {"installer":{"name":"uv","version":"0.12.12","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":null,"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":null}
|
Security for chujie-llm-toolkit 0.1.0
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.