chore: 增加issue自动回复

Author: tangying1027

.github/workflows/issue-automated.yml

@@ -0,0 +1,56 @@
+name: Issue Automated Processing
+
+on:
+  issues:
+    types: [opened, reopened, edited]
+  workflow_dispatch:
+    inputs:
+      issue_number:
+        description: 'Issue number to process'
+        required: true
+        type: number
+
+permissions:
+  issues: write
+  pull-requests: read
+  contents: read
+
+jobs:
+  issue-response:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v3
+
+      - name: Set up Node.js
+        uses: actions/setup-node@v3
+        with:
+          node-version: 20
+
+      - name: Install dependencies
+        run: npm install openai @antv/mcp-server-antv
+          
+      - name: Process Issue
+        uses: actions/github-script@v7
+        with:
+          github-token: ${{ secrets.GITHUB_TOKEN }}
+          script: |
+            let issue = context.payload.issue;
+            
+            if (context.eventName === 'workflow_dispatch') {
+              const issueNumber = context.payload.inputs.issue_number;
+              const { data: issueData } = await github.rest.issues.get({
+                owner: context.repo.owner,
+                repo: context.repo.repo,
+                issue_number: issueNumber
+              });
+              issue = issueData;
+            }
+            
+            const script = require('./.github/workflows/scripts/issue-automated.js');
+            await script({
+              github,
+              core,
+              context,
+              issue
+            });

.github/workflows/scripts/issue-automated.js

@@ -0,0 +1,126 @@
+const { OpenAI } = require("openai");
+const { QueryAntVDocumentTool, ExtractAntVTopicTool }=  require('@antv/mcp-server-antv/build/tools');
+
+
+/**
+ * @param {Object} param
+ * @param {import('@actions/github').GitHub} param.github
+ * @param {import('@actions/core')} param.core
+ * @param {Object} param.context GitHub Action context
+ */
+module.exports = async ({ github, core, context, issue }) => {
+  try {
+    core.info('开始处理 issue...');
+    const library = `${context.repo.repo}`
+    if (!issue) {
+      core.setFailed('找不到 issue 信息');
+      return;
+    }
+    
+    const issueNumber = issue.number;
+    const issueTitle = issue.title;
+    
+    core.info(`处理 issue #${issueNumber}: ${issueTitle}`);
+    
+    const combinedQuery = prepareAIPrompt(context, issue);
+    
+   
+    const topicExtractionResult = await ExtractAntVTopicTool.run({ query: combinedQuery });
+   
+    const aiResponse = await getAIResponse(core,  topicExtractionResult.content[0].text);
+    const jsonMatch = aiResponse.match(/```json\s*(\{[\s\S]*?\})\s*```/);
+    const processedTopicContent = JSON.parse(jsonMatch[1]);
+
+    const queryDocumentParams = {
+        library,
+        query: combinedQuery,
+        topic: processedTopicContent.topic,
+        intent: processedTopicContent.intent,
+        tokens: 5000,
+        ...(processedTopicContent.subTasks && { subTasks: processedTopicContent.subTasks }),
+      };
+
+      const documentationResult = await QueryAntVDocumentTool.run(queryDocumentParams);
+
+      const response = await getAIResponse(core,  documentationResult.content[0].text);
+
+        await github.rest.issues.createComment({
+            issue_number: issue.number,
+            owner: context.repo.owner,
+            repo: context.repo.repo,
+            body: `@${issue.user.login} 您好！以下是关于您问题的自动回复：\n\n${response}\n\n---\n*此回复由 AI 助手自动生成。如有任何问题，我们的团队会尽快跟进。*`
+        });
+    
+    core.info('Issue 处理完成');
+    
+  } catch (error) {
+    core.setFailed(`处理 issue 失败: ${error.message}`);
+    core.error(error.stack);
+  }
+};
+
+function prepareAIPrompt(context, issue) {
+    return `
+    你是 ${context.repo.repo} 项目的智能助手。这是一个处理 GitHub issue 的自动回复系统。
+    请分析以下 issue 并提供专业、有帮助的回复。
+
+    ## 当前 Issue
+    - 标题: ${issue.title}
+    - 内容: ${issue.body}
+
+    请提供完整、有帮助的回复，但不要过于冗长。回复应该条理清晰，使用适当的 Markdown 格式。
+`;
+}
+
+/**
+ * 调用 GitHub AI API 获取回复
+ */
+async function getAIResponse(core, userQuestion) {
+  try {
+    core.info('正在调用 GitHub AI API...');
+    
+    const token = process.env.GH_TOKEN;
+    
+    if (!token) {
+      throw new Error('未找到 GH_TOKEN 环境变量');
+    }
+    
+    const endpoint = "https://models.github.ai/inference";
+    const model = "openai/gpt-4.1";
+    
+    const client = new OpenAI({ 
+      baseURL: endpoint, 
+      apiKey: token 
+    });
+
+    const response = await client.chat.completions.create({
+      messages: [
+        { role: "user", content: userQuestion }
+      ],
+      temperature: 0.7,
+      top_p: 1.0,
+      model: model
+    });
+
+    core.info('成功获取 AI 响应');
+    return response.choices[0].message.content;
+    
+  } catch (error) {
+    core.warning(`调用 GitHub AI API 失败: ${error.message}`);
+    // 默认回复
+    return `
+    感谢您提交这个 issue！
+
+    我们的团队会尽快查看您的问题。为了帮助我们更快地解决，请确保您提供了:
+
+    - 问题的详细描述
+    - 复现步骤 (如果是 bug)
+    - 预期行为和实际行为
+    - 使用的版本信息
+
+    谢谢您的理解与支持！
+`;
+  }
+}
+
+

package.json

@@ -65,13 +65,13 @@
     "signal-exit": "3.0.3"
   },
   "devDependencies": {
-    "babel-plugin-search-and-replace": "^1.1.1",
     "@octokit/core": "^4.2.0",
     "@types/jest": "^26.0.20",
     "@types/jest-image-snapshot": "^6.1.0",
     "@typescript-eslint/eslint-plugin": "^5.10.0",
     "@typescript-eslint/parser": "^5.10.0",
     "adm-zip": "^0.5.10",
+    "babel-plugin-search-and-replace": "^1.1.1",
     "canvas-to-buffer": "^1.1.1",
     "cross-env": "^6.0.3",
     "danger": "^11.2.3",