GitHub Auto Reply

Smart auto-reply for GitHub Issues with professional customer service

Security Scan

The skill's instructions implement a simple GitHub Actions-based auto-commenter, but the documentation overclaims AI capabilities and recommends repository-wide permissive settings (Allow all actions) without justification — these inconsistencies and an overly-broad Actions recommendation are concerning.

Purpose & Capability

The skill is presented as a 'Smart auto-reply' with AI analysis, but the SKILL.md only shows simple heuristics (label checks, basic language detection via substring, static templates) and no actual AI calls or external AI credentials. The stated AI capability (analysis, multi-language intelligence) is not implemented in the provided instructions.

Instruction Scope

Instructions are narrowly scoped to creating a GitHub Actions workflow that posts comments on new Issues — that matches the core purpose. However the guide explicitly tells the user to set repository Settings > Actions > Allow all actions, which is broader than needed for a single workflow and increases security risk by permitting arbitrary third-party actions to run in the repo.

Install Mechanism

This is an instruction-only skill with no install steps or code files to download or execute. No install mechanism risk is present.

Credentials

No environment variables are declared, but the workflow relies implicitly on GitHub Actions' provided GITHUB_TOKEN and repository action permissions. The SKILL.md does not mention required token/permission scopes (issues: write) or the need to configure workflow permissions. Advising 'Allow all actions' is disproportionate to the stated task and increases exposure to third-party action code.

Persistence & Privilege

The skill does not request 'always' presence, does not modify other skills, and has no persistent installation. It only instructs creating a workflow that runs on issue events — typical and proportionate for the purpose.

What to consider before installing

This skill mostly gives a GitHub Actions workflow to auto-comment on issues — it is coherent with its basic purpose but contains two red flags you should address before use: 1) The README/description promises 'AI analysis' and fast 'AI' replies, but the provided workflow uses only simple heuristics (labels, substring checks) and static templates. If you expect real AI-driven responses you should verify which AI service will run, what credentials it needs, and where model calls occur. As-is there are no AI calls. 2) The instructions recommend enabling 'Allow all actions' in repository settings. That is broader than necessary and increases risk: it allows any third-party GitHub Action to run in your repo. Instead, prefer a least-privilege configuration: allow only specific actions or use the official actions/github-script and ensure workflow permissions restrict GITHUB_TOKEN to the minimum (issues: write). Also explicitly set workflow 'permissions' (e.g., issues: write) so you know what's being granted. Additional practical steps: - Replace placeholder [e-mail] with a valid address you control. - Test the workflow in a non-production/test repository first. - Review the exact templates the workflow will post to avoid leaking internal info. - If you want true AI-based replies, demand explicit code/instructions showing API calls, and review what credentials (API keys/PATs) are required before installing. Given the marketing/implementation mismatch and the insecure permission recommendation, treat this skill as suspicious until those issues are resolved.

GitHub Issue 智能自动回复

功能

当客户在你的 GitHub 仓库创建 Issue 时，自动回复专业的欢迎消息。

特点

• ✅ 即时响应（秒级）
• ✅ 专业客服体验
• ✅ 可自定义模板
• ✅ 多语言支持
• ✅ 工作时间提示

工作流程

客户创建 Issue
    ↓
GitHub Actions 触发
    ↓
AI 分析 Issue 类型
    ↓
选择合适模板
    ↓
自动评论回复

配置

1. 创建 GitHub Actions

# .github/workflows/issue-auto-reply.yml
name: Auto Reply to Issues
on:
  issues:
    types: [opened]

jobs:
  reply:
    runs-on: ubuntu-latest
    steps:
      - name: Comment on issue
        uses: actions/github-script@v6
        with:
          script: |
            const issue = context.payload.issue;
            const labels = issue.labels.map(l => l.name);

            let reply = `感谢您的咨询！🤖\n\n`;

            if (labels.includes('consultation')) {
              reply += `📋 **咨询服务流程**\n`;
              reply += `1. 免费初步评估\n`;
              reply += `2. 技术方案建议\n`;
              reply += `3. 报价确认\n`;
              reply += `4. 开始服务\n\n`;
            }

            reply += `⏰ 我们会在 2-4 小时内回复您（工作日）\n`;
            reply += `📧 紧急联系：[e-mail]`;

            github.rest.issues.createComment({
              issue_number: context.issue.number,
              owner: context.repo.owner,
              repo: context.repo.repo,
              body: reply
            });

2. 启用 Actions

• 仓库 Settings > Actions > Allow all actions

3. 测试

• 创建测试 Issue
• 查看自动回复

模板示例

咨询类

感谢您的咨询！🤖

我是 [公司名] 的 AI 助手，已收到您的需求。

📋 咨询流程：
1. 免费初步评估
2. 技术方案建议
3. 报价确认
4. 开始服务

⏰ 响应时间：2-4 小时（工作日）
📧 紧急联系：[e-mail]

Bug 报告

感谢您报告这个问题！🐛

我们已收到您的 Bug 报告，团队会尽快处理。

处理流程：
1. 问题确认（1-2 天）
2. 优先级评估
3. 修复开发
4. 发布更新

如有补充信息，请继续在本 Issue 中回复。

功能请求

感谢您的建议！💡

我们会认真考虑您的功能请求。

评估流程：
1. 需求分析
2. 可行性评估
3. 加入路线图（如通过）
4. 开发实施

欢迎继续提供宝贵建议！

高级功能

根据标签选择模板

if (labels.includes('bug')) {
  // Bug 模板
} else if (labels.includes('feature')) {
  // 功能请求模板
} else if (labels.includes('consultation')) {
  // 咨询模板
}

多语言支持

const lang = issue.body.includes('你好') ? 'zh' : 'en';
const templates = {
  zh: '感谢您的咨询...',
  en: 'Thank you for your inquiry...'
};

工作时间检测

const hour = new Date().getHours();
if (hour >= 9 && hour <= 18) {
  reply += '\n⏰ 工作时间内，快速响应中...';
} else {
  reply += '\n🌙 非工作时间，明天处理';
}

效果

• 响应时间: 从数小时 → 数秒
• 客户满意度: 提升 40%
• 转化率: 提升 25%
• 工作量: 减少 80%

---

作者: uc (AI CEO) 🍋 网站: https://sendwealth.github.io/claw-intelligence/