Skillget

aiway

mcp
Guvenlik Denetimi
Gecti
Health Gecti
  • License — License: Apache-2.0
  • Description — Repository has a description
  • Active repo — Last push 0 days ago
  • Community trust — 12 GitHub stars
Code Gecti
  • Code scan — Scanned 12 files during light audit, no dangerous patterns found
Permissions Gecti
  • Permissions — No dangerous permissions requested
Purpose
This tool is a lightweight, high-performance API and AI gateway built in Rust. It provides dynamic routing, request forwarding, and management for backend services, with dedicated support for Model Context Protocol (MCP) integration.

Security Assessment
The overall risk is rated as Low. As an API gateway, this tool's core function is to receive and proxy network requests, which is expected behavior and not a malicious flag. A light code scan of 12 files found no dangerous patterns, hardcoded secrets, or requests for overly broad system permissions. Its memory management is inherently secured by the Rust language. However, administrators should note that the default login credentials for the management console are `admin` / `admin`, which must be changed immediately upon deployment to prevent unauthorized access.

Quality Assessment
The project is actively maintained, with its most recent code push occurring today. It uses the standard Apache-2.0 license, making it highly accessible for open-source and commercial use. Community trust is currently in its early stages, reflected by a modest 12 GitHub stars. Despite the small community footprint, the project demonstrates good engineering quality through its quick build process, multi-architecture support, and solid performance benchmarks.

Verdict
Safe to use, provided you change the default administrative credentials before exposing the console to any network.
SUMMARY

API + AI 融合网关

README.md

aiway (API and AI Gateway) 是一个基于 Rust 开发的轻量、高性能 API + AI 网关，致力于提供稳定、高效、可扩展的请求转发与管理解决方案。

凭借 Rust 的内存安全特性和零成本抽象优势，aiway 在保证高性能的同时，提供了卓越的安全性和稳定性。

与其他网关不同，aiway无需复杂的配置和环境依赖，网关节点可动态扩容/缩容，轻松应对百万并发连接。

支持的系统

  • Linux (x86_64 / arm64)
  • macOS (arm64)
  • openEuler (x86_64)
  • UOS Server (x86_64)
  • KylinOS (x86_64)

支持的协议

  • HTTP（IPv4/IPv6）
  • SSE
  • WebSocket
  • MCP

文档

详细文档请访问：https://aiway.coderbox.cn/doc.html

快速开始

方式一：使用预编译版本

# 下载并解压
curl -L https://github.com/xgpxg/aiway/releases/latest/download/aiway-linux-amd64-standalone.tar.gz | tar -zxvf - -C .

# 启动服务
./aiway

注意：预编译版本基于 glibc 2.34 构建，如果您的系统glibc版本低于 2.34，请从源码构建。

方式二：从源码构建

# 构建 Gateway
cargo build --bin gateway

# 构建 Console
cargo build --bin console

# 构建 Logg
cargo build --bin logg

# 运行
cargo run --bin aiway

访问服务

启动参数

./aiway -h

Usage: aiway [OPTIONS]

Options:
  -a, --address <ADDRESS>            Listen address, like 127.0.0.1 [default: 127.0.0.1]
  -p, --port <PORT>                  Port [default: 7000]
      --gateway-port <GATEWAY_PORT>  Gateway port [default: 7001]
  -h, --help                         Print help
  -V, --version                      Print version

核心功能

  • 动态路由 - 灵活配置请求路由规则
  • 服务管理 - 统一管理后端服务实例
  • 插件系统 - 可扩展的插件架构
  • 安全防护 - 内置防火墙和安全验证机制
  • API Key 管理 - 统一的 API 密钥管理
  • 日志监控 - 完整的日志存储和实时监控
  • 可视化面板 - 直观的管理控制台
  • AI 模型代理 - 智能 AI 模型代理和请求转发
  • MCP 集成 - Model Context Protocol 支持

插件生态

官方插件

我们提供了一系列常用插件，访问 aiway-plugins 获取更多信息。

自定义插件

如需开发自定义插件，请参考 插件开发文档。

界面预览

Dashboard

性能

  • Ubuntu 24.04, Intel i7-12700K, 12 Cores, 16 GB RAM
wrk http://127.0.0.1:7001/api/hello -t 12 -c 100

Running 10s test @ http://127.0.0.1:7001/api/hello
  12 threads and 100 connections
  Thread Stats   Avg      Stdev     Max   +/- Stdev
    Latency     0.89ms  780.09us  28.66ms   95.87%
    Req/Sec     9.54k     1.12k   18.36k    93.70%
  1144556 requests in 10.10s, 287.07MB read
Requests/sec: 113330.59
Transfer/sec:     28.43MB
  • Ubuntu 22.04, 4 Cores, 8 GB RAM
wrk http://127.0.0.1:7001/hello -t 4 -c 100

Running 10s test @ http://127.0.0.1:7001/hello
  4 threads and 100 connections
  Thread Stats   Avg      Stdev     Max   +/- Stdev
    Latency     6.70ms    3.05ms  36.85ms   73.84%
    Req/Sec     7.61k   770.50     9.22k    64.50%
  302860 requests in 10.01s, 75.96MB read
Requests/sec:  30265.20
Transfer/sec:      7.59MB

以下为在老版本上的测试数据（<0.2.2），新版本会快30%~40%

  • openEuler 24.03 (LTS-SP3), Intel(R) Xeon(R) 6982P-C, 4 Cores, 8 GB RAM
wrk http://127.0.0.1:7001/hello -t 4 -c 100

Running 10s test @ http://127.0.0.1:7001/hello
  4 threads and 100 connections
  Thread Stats   Avg      Stdev     Max   +/- Stdev
    Latency     3.80ms    1.22ms  13.60ms   70.35%
    Req/Sec     6.62k   380.06     7.43k    62.75%
  263639 requests in 10.01s, 74.67MB read
Requests/sec:  26332.11
Transfer/sec:      7.46MB
  • UOS Server 20, Intel(R) Xeon(R) Platinum, 4 Cores, 8 GB RAM
wrk http://127.0.0.1:7001/hello -t 4 -c 100

Running 10s test @ http://127.0.0.1:7001/hello
  8 threads and 100 connections
  Thread Stats   Avg      Stdev     Max   +/- Stdev
    Latency     3.61ms    2.48ms  84.71ms   93.78%
    Req/Sec     3.47k   452.12    11.40k    83.88%
  276965 requests in 10.06s, 78.45MB read
Requests/sec:  27518.91
Transfer/sec:      7.79MB

贡献

欢迎提交 Issue 和 Pull Request 来帮助改进 aiway！

Yorumlar (0)

Sonuc bulunamadi