Token导航 LogoToken导航TokenDH.com
前端设计敏感数据github未标认证来源可访问许可证需确认审计通过

custom-test-process定制测试流程

Agent Skill

用于辅助测试设计、自动化测试、用例整理和回归验证。它适合让 Agent 编写单元测试、端到端测试、测试计划或根据失败日志定位问题。使用时需要确认项目测试框架、运行命令和夹具数据,避免为了通过测试而改坏真实逻辑;涉及浏览器或外部服务时,应区分本地模拟、测试环境和生产环境。

总安装

514

周安装

21

GitHub Stars

12

下载量

166
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

unknown

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:custom-test-process(定制测试流程)
来源仓库:https://github.com/edanstarfire/claudecode_webui
仓库路径:skills/custom-test-process
安装命令:
npx skills add https://github.com/edanstarfire/claudecode_webui --skill custom-test-process
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/edanstarfire/claudecode_webui --skill custom-test-process

简介

custom-test-process 运行完整测试周期,包括启动服务与验证健康端点。

  • 依赖 custom-environment-setup 提供的端口信息避免冲突。
  • 覆盖前后端接口测试,确保改动未破坏现有功能。
  • 测试失败时应区分预期行为与真实缺陷,避免误报干扰。
  • 适用宿主包括 Codex、Claude、Cursor、Gemini CLI,接入前应确认版本、权限和运行环境要求。

SKILL.md

Custom Test Process

Purpose

This is a project-specific custom skill called by the Builder workflow to run the full test cycle. It handles starting servers, running tests, and verifying health endpoints specific to this project (claudecode_webui).

Generic workflow skills invoke this skill if it exists; if absent, the test step is skipped.

When Called

The Builder invokes this skill from its working directory (the worktree). Environment variables (ports) come from custom-environment-setup.

Input

Environment from custom-environment-setup:

  • BACKEND_PORT: Backend server port (8000 + issue_number % 1000)
  • VITE_PORT: Vite dev server port (5000 + issue_number % 1000)
  • TEST_AUTH_TOKEN: Fixed auth token for test servers (default: test)

- Pinned so the token survives restarts during testing - Included in all URLs reported to the user

Test Lifecycle

This skill owns the full test lifecycle in a single invocation:

1. Start Backend Server

CRITICAL: Unset the CLAUDECODE environment variable before starting the backend. The Claude Agent SDK includes an undocumented safety check that prevents it from running inside another Claude Code instance. Since the builder agent runs inside Claude Code, this env var is inherited by child processes. Our application launches its own Claude Code SDK instances, so if CLAUDECODE is set, those SDK sessions will halt prematurely.

env -u CLAUDECODE uv run python main.py --host 0.0.0.0 --debug-all --port ${BACKEND_PORT} --token ${TEST_AUTH_TOKEN:-test} &

Wait for server to be ready:

# Wait up to 30 seconds for backend
for i in $(seq 1 30); do
    curl -s http://localhost:${BACKEND_PORT}/health > /dev/null 2>&1 && break
    sleep 1
done

2. Start Frontend Dev Server (if frontend changed)

CRITICAL: Set VITE_BACKEND_PORT so the vite dev server proxies to the correct per-issue backend port instead of the default 8001:

cd frontend && VITE_BACKEND_PORT=${BACKEND_PORT} npm run dev -- --port ${VITE_PORT} &

The VITE_BACKEND_PORT environment variable is read by vite.config.js to configure the proxy target. Without this, the frontend would proxy requests to port 8001 (the default) instead of the issue-specific backend port.

3. Run Unit Tests

uv run pytest src/tests/ -v

4. Verify Health Endpoints

# Check backend health
curl -s http://localhost:${BACKEND_PORT}/health

# Check frontend (if started)
curl -s http://localhost:${VITE_PORT}

5. Leave Servers Running

CRITICAL: Do NOT stop servers after testing. Leave them running for user review. The custom-cleanup-process skill handles stopping servers later.

6. Report Server URLs to User

When reporting that servers are running, include the auth token in the URLs:

Backend:  http://localhost:${BACKEND_PORT}/?token=${TEST_AUTH_TOKEN:-test}
Frontend: http://localhost:${VITE_PORT}/?token=${TEST_AUTH_TOKEN:-test}

This lets the user click the URL and authenticate automatically.

Test Verification

  • Server starts without errors
  • Health endpoint returns success
  • Unit tests pass
  • No regressions introduced

Usage by Generic Skills

The Builder workflow calls this skill like:

Invoke custom-test-process skill

The skill uses port information from the environment (provided by custom-environment-setup). It may use the process-manager skill internally for process management. If this skill does not exist, the generic workflow skips the test step.

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

Codex

36.84%
按下载量换算61

Claude

30.99%
按下载量换算51

Cursor

19.77%
按下载量换算33

Gemini CLI

9.67%
按下载量换算16

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

敏感数据

该 Skill 可能接触密钥、Token、环境变量或敏感配置,应进入高风险复核队列,默认不自动发布。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills