搜索

PythonReactSEOPDF
GitHub Skill

integration-tests

Use this skill when writing or running integration tests. Provides approved model names and integration test guidelines.

vibelang-org☆ 0 仓库 Stars0 仓库 Forks

安装与使用

请根据来源仓库和你的 AI 工具说明安装此技能。先阅读技能文件和依赖要求,再添加到工作流。

查看来源 ↗使用指南

技能说明

Integration Test Guidelines

Guidelines for writing and running integration tests that make real API calls.

Approved Models

Use these models for integration tests:

Provider Model Name API Key Env Var
Anthropic claude-haiku-4-5-20251001 ANTHROPIC_API_KEY
OpenAI gpt-5-mini OPENAI_API_KEY
Google gemini-3-flash GOOGLE_API_KEY
Why These Models?
  • Fast: Optimized for low latency
  • Cost-effective: Cheapest tier for each provider
  • Capable: Still highly capable for test scenarios

Writing Integration Tests

Example Vibe Code
// Anthropic
model anthropic = { name: "claude-haiku-4-5-20251001", apiKey: env("ANTHROPIC_API_KEY") }

// OpenAI
model openai = { name: "gpt-5-mini", apiKey: env("OPENAI_API_KEY") }

// Google
model google = { name: "gemini-3-flash", apiKey: env("GOOGLE_API_KEY") }
Test File Location

Integration tests live in: tests/integration/

Running Integration Tests
# Run all integration tests (costs money, ~40s)
bun run test:integration

# Run a single integration test
bun test tests/integration/specific-test.test.ts

Important Notes

  1. Do NOT run full integration suite routinely - it costs money
  2. Use unit tests first - bun run test is fast and free
  3. Run targeted integration tests when verifying specific provider behavior
  4. Update model names when providers release new versions

使用示例

查看来源仓库中的使用示例与配置说明。

查看来源 ↗

来源历史

收录时间
2026-01-20
索引更新时间
2026-01-26
Branch
main

收录与同步日期属于本站记录。源码修改历史请查看来源仓库。

GitHub ↗

chat Comments (0)

chat_bubble_outline

No comments yet. Be the first to share your thoughts!