工具简介
Keploy 是一款面向「开发者工具与 API」场景的 AI 工具，价格信息以官网为准，主要通过网页端提供。官网域名是 keploy.io。
如何使用
- 访问 keploy.io 完成注册，先阅读快速开始文档
- 建立测试项目，按示例跑通第一个用例
- 根据实际需求调整参数与调用方式
- 确认稳定后接入生产流程，并做好错误处理与限流
查看官网原文（英文）
以下为收录时的原文，语言与页面不同，未做翻译。
Open source Tool for converting user traffic to Test Cases and Data Stubs.
官网介绍
Keploy builds a Digital Twin of your production behavior — real API traffic, mocks, and dependencies — so every AI- or human-written change is verified in CI before it ships. Open source, zero code changes.
官方常见问题
- What is Keploy?
- Keploy is an open-source, AI-powered testing platform that turns real application traffic into ready-to-run tests and dependency mocks. It records incoming API requests and outgoing calls to databases, queues, and external services, then replays them as deterministic tests locally, in CI, or across Kubernetes.
- Is Keploy an API testing or integration testing tool?
- Keploy supports both. Its AI testing platform generates API tests from specifications, requests, and application context. Its record-and-replay platform creates integration and regression tests from real traffic while virtualizing databases, queues, APIs, and other dependencies.
- How does Keploy work?
- Keploy uses eBPF and a transparent network proxy to capture incoming requests and outgoing dependency calls without instrumenting application code. It stores the captured behavior as editable YAML tests and mocks, then replays the requests against a new build while serving recorded dependency responses.
- Does Keploy require code changes?
- No application code changes are required for traffic-based record and replay. Keploy observes network activity using eBPF and proxy-based interception, so developers can run their applications normally while Keploy captures requests, database calls, API calls, and other supported dependencies.
- How does generating tests from real traffic improve coverage?
- Real traffic gives Keploy the requests, responses, dependency interactions, and data states that the application actually handles. Keploy converts this behavior into reproducible tests and mocks, then uses API schemas and application context to generate additional edge cases and flows beyond the recorded examples.
- How is Keploy different from Postman?
- Postman is primarily used to design, send, organize, and manually test API requests. Keploy can automatically generate tests from real application traffic and capture the backend dependencies involved, including databases, queues, and external APIs. This lets teams replay complete application behavior without manually recreating every request, mock, or dependency state.
编程与调试 · 常用提示词
下面这些都能直接复制，把里面的变量换成你自己的情况就能用。
- 数据迁移与回滚：先想好怎么退回去为一次表结构变更设计向前迁移、向后回滚、双写过渡与校验方案。
- 写一个不会误删文件的 Shell 脚本按「参数校验—干跑—确认—执行—日志」的骨架写脚本，重点处理空格路径、失败退出与幂等。
- TypeScript 类型设计：让错误在编译期就暴露把「什么都能传」的 interface 改成有约束的类型，重点讲可辨识联合与窄化。