🤖 AI Bugfix Agent 完整技术实施方案

# Bug 事件标准化数据模型 from pydantic import BaseModel, Field from datetime import datetime from typing import List, Dict, Any, Optional from enum import Enum class BugSeverity(Enum): CRITICAL = "critical" HIGH = "high" MEDIUM = "medium" LOW = "low" class BugSource(Enum): GITHUB = "github" JIRA = "jira" EMAIL = "email" SLACK = "slack" MONITORING = "monitoring" class BugEvent(BaseModel): """标准化 Bug 事件模型""" id: str = Field(..., description="唯一标识符") source: BugSource title: str description: str severity: BugSeverity priority: int = Field(ge=1, le=5) status: str = "new" # 复现信息 reproduction_steps: List[str] = [] environment: Dict[str, str] = {} # 代码定位 file_path: Optional[str] = None line_number: Optional[int] = None commit_hash: Optional[str] = None # 附件与日志 attachments: List[str] = [] error_logs: str = "" stack_trace: Optional[str] = None # 元数据 created_at: datetime updated_at: datetime reporter: str assignee: Optional[str] = None # 扩展字段 metadata: Dict[str, Any] = {} class Config: json_schema_extra = { "example": { "id": "BUG-2026-001", "source": "github", "title": "SQL 注入漏洞 - 用户登录接口", "severity": "critical", "priority": 1 } }

# 基于语义相似度的 Bug 去重算法 from sentence_transformers import SentenceTransformer from sklearn.metrics.pairwise import cosine_similarity import numpy as np class BugDeduplicator: def __init__(self): # 使用预训练模型进行语义编码 self.model = SentenceTransformer('all-MiniLM-L6-v2') self.bug_embeddings = {} self.similarity_threshold = 0.85 def add_bug(self, bug_id: str, title: str, description: str): """添加 Bug 并计算嵌入向量""" text = f"{title} {description}" embedding = self.model.encode([text])[0] self.bug_embeddings[bug_id] = embedding def find_duplicates(self, new_bug: BugEvent) -> List[str]: """查找相似的 Bug""" new_text = f"{new_bug.title} {new_bug.description}" new_embedding = self.model.encode([new_text])[0] duplicates = [] for bug_id, embedding in self.bug_embeddings.items(): similarity = cosine_similarity( [new_embedding], [embedding] )[0][0] if similarity >= self.similarity_threshold: duplicates.append((bug_id, similarity)) # 按相似度排序 duplicates.sort(key=lambda x: x[1], reverse=True) return [bug_id for bug_id, _ in duplicates] def merge_bugs(self, primary_id: str, duplicate_ids: List[str]): """合并重复 Bug，聚合多渠道信息""" # 实现逻辑：保留主 Bug，附加渠道信息 pass

// Agent Trace 记录格式（JSON Schema） { "$schema": "https://agenttrace.io/schema/v1.json", "trace_id": "550e8400-e29b-41d4-a716-446655440000", "session_id": "session-abc-123", "timestamp": "2026-03-03T14:30:00Z", // 贡献者类型 "contributor_type": "ai", "contributor_subtype": "claude-3.5-sonnet", // 代码范围 "code_range": { "file": "src/auth/login.py", "start_line": 45, "end_line": 78, "content_hash": "sha256:a1b2c3d4e5f6...", "function_name": "authenticate_user" }, // 对话上下文 "conversation": { "prompt": "修复 SQL 注入漏洞，使用参数化查询", "dialogue_id": "dialog-xyz-789", "turn_number": 3 }, // 模型信息 "model_info": { "provider": "anthropic", "model": "claude-3.5-sonnet-20260101", "temperature": 0.3, "max_tokens": 4096 }, // 存储后端 "storage": { "backend": "git_notes", "ref": "refs/notes/agent-traces", "commit": "abc123def456..." }, // 扩展元数据 "extensions": { "security_review": "passed", "test_coverage": 0.92 } }

#!/usr/bin/env python3 # Agent Trace Git 集成工具 import subprocess import json import hashlib from pathlib import Path class AgentTraceGit: def __init__(self, repo_path: str): self.repo_path = Path(repo_path) self.notes_ref = "refs/notes/agent-traces" def compute_content_hash(self, file_path: str, start_line: int, end_line: int) -> str: """计算代码片段的内容哈希""" full_path = self.repo_path / file_path with open(full_path, 'r', encoding='utf-8') as f: lines = f.readlines() snippet = ''.join(lines[start_line-1:end_line]) return hashlib.sha256(snippet.encode()).hexdigest() def add_trace(self, trace_data: dict): """添加 Agent Trace 记录到 Git Notes""" # 计算内容哈希 content_hash = self.compute_content_hash( trace_data['code_range']['file'], trace_data['code_range']['start_line'], trace_data['code_range']['end_line'] ) # 生成 Git Notes 键 note_key = f"trace:{content_hash}" # 添加 Git Note cmd = [ 'git', 'notes', '--ref', self.notes_ref, 'add', '-f', # 强制覆盖 '-m', json.dumps(trace_data), 'HEAD' ] result = subprocess.run( cmd, cwd=self.repo_path, capture_output=True, text=True ) if result.returncode != 0: raise Exception(f"Git notes add failed: {result.stderr}") return content_hash def query_traces(self, file_path: str = None, commit_hash: str = None) -> list: """查询 Agent Trace 记录""" cmd = [ 'git', 'notes', '--ref', self.notes_ref, 'show' ] if commit_hash: cmd.append(commit_hash) result = subprocess.run( cmd, cwd=self.repo_path, capture_output=True, text=True ) if result.returncode != 0: return [] # 解析 Notes 输出 traces = [] for line in result.stdout.splitlines(): try: trace = json.loads(line) if file_path is None or trace['code_range']['file'] == file_path: traces.append(trace) except json.JSONDecodeError: continue return traces def enhanced_blame(self, file_path: str, line_number: int) -> dict: """增强的 Git Blame，包含 AI 贡献信息""" # 执行标准 git blame cmd = [ 'git', 'blame', '-L', f"{line_number},{line_number}", '--porcelain', file_path ] result = subprocess.run( cmd, cwd=self.repo_path, capture_output=True, text=True ) blame_info = self._parse_blame(result.stdout) # 查询对应的 Agent Trace commit_hash = blame_info.get('commit') traces = self.query_traces(file_path, commit_hash) # 合并信息 blame_info['agent_trace'] = traces return blame_info def _parse_blame(self, blame_output: str) -> dict: """解析 Git Blame 输出""" info = {} for line in blame_output.splitlines(): if line.startswith("author "): info['author'] = line[7:] elif line.startswith("author-mail "): info['email'] = line[12:] elif line.startswith("author-time "): info['timestamp'] = int(line[12:]) return info

#!/usr/bin/env python3 # Claude Code 修复生成引擎 import anthropic import json from typing import List, Dict, Optional from pathlib import Path class ClaudeCodeFixEngine: def __init__(self, api_key: str, model: str = "claude-3-5-sonnet-20260101"): self.client = anthropic.Anthropic(api_key=api_key) self.model = model self.system_prompt = self._load_system_prompt() def _load_system_prompt(self) -> str: """加载系统提示词（CLAUDE.md 规范）""" return """ 你是一位资深软件工程师和安全专家，负责自动修复代码中的 Bug 和安全漏洞。 ## 强制检查清单（必须执行）： 1. ✅ 扫描硬编码密钥和敏感信息 2. ✅ 检查 SQL 注入、XSS、CSRF 等安全漏洞 3. ✅ 验证所有用户输入 4. ✅ 确保错误处理完善 5. ✅ 检查类型安全和空指针问题 6. ✅ 运行相关测试用例 ## 修复原则： - 最小改动原则：只修改必要代码 - 向后兼容：不破坏现有 API - 代码规范：遵循项目编码风格 - 测试覆盖：修复必须包含测试 ## 禁止操作： ❌ 修改系统文件 ❌ 访问敏感目录（/etc, /root 等） ❌ 执行未授权的网络请求 ❌ 引入新的外部依赖（未经审批） ## 输出格式： 1. 问题分析 2. 修复方案说明 3. 修复后的代码（完整函数/类） 4. 测试用例建议 5. 影响范围评估 """ def analyze_bug(self, bug_event: dict, code_context: str) -> dict: """分析 Bug 并生成修复方案""" prompt = f""" ## Bug 信息 - 标题：{bug_event['title']} - 严重程度：{bug_event['severity']} - 描述：{bug_event['description']} ## 错误堆栈 {bug_event.get('stack_trace', 'N/A')} ## 相关代码 ```python {code_context} ``` ## 任务 1. 分析 Bug 根本原因 2. 提供详细的修复方案 3. 生成修复后的完整代码 4. 编写测试用例验证修复 5. 评估影响范围 请按照强制检查清单逐项检查，确保修复安全有效。 """ response = self.client.messages.create( model=self.model, max_tokens=4096, system=self.system_prompt, messages=[ {"role": "user", "content": prompt} ], temperature=0.3 ) return self._parse_fix_response(response.content[0].text) def generate_security_tests(self, code: str, vulnerability_type: str) -> List[str]: """生成针对性安全测试用例""" prompt = f""" 针对以下代码的 {vulnerability_type} 漏洞，生成 20 个破坏性测试用例： ```python {code} ``` 测试用例应覆盖： - 边界条件（null、空字符串、负数、超大值） - SQL 注入 payload - XSS 攻击向量 - 路径遍历尝试 - 认证绕过尝试 每个测试用例包含： 1. 测试描述 2. 输入数据 3. 预期结果 """ response = self.client.messages.create( model=self.model, max_tokens=4096, messages=[{"role": "user", "content": prompt}] ) return self._parse_test_cases(response.content[0].text) def _parse_fix_response(self, response_text: str) -> dict: """解析修复响应""" # 使用正则表达式提取各部分 import re fix = { "analysis": "", "solution": "", "fixed_code": "", "tests": [], "impact": "" } # 提取代码块 code_pattern = rr'```(?:python)?\s*(.*?)```' code_matches = re.findall(code_pattern, response_text, re.DOTALL) if code_matches: fix['fixed_code'] = code_matches[-1] # 最后一个代码块 fix['raw_response'] = response_text return fix

// Jenkinsfile - AI Bugfix 完整流水线 pipeline { agent { kubernetes { yaml """ apiVersion: v1 kind: Pod spec: containers: - name: bugfix-agent image: bugfix-agent:latest command: ['cat'] tty: true resources: limits: cpu: '2' memory: 4Gi requests: cpu: '500m' memory: 1Gi - name: docker image: docker:24.0 command: ['cat'] tty: true volumeMounts: - name: docker-sock mountPath: /var/run/docker.sock volumes: - name: docker-sock hostPath: path: /var/run/docker.sock """ } } environment { BUG_ID = params.BUG_ID REPO_URL = params.REPO_URL ANTHROPIC_API_KEY = credentials('anthropic-api-key') GITHUB_TOKEN = credentials('github-token') SONAR_TOKEN = credentials('sonar-token') SNYK_TOKEN = credentials('snyk-token') } parameters { string(name: 'BUG_ID', defaultValue: '', description: 'Bug ID') string(name: 'REPO_URL', defaultValue: '', description: '代码仓库 URL') string(name: 'BRANCH', defaultValue: 'main', description: '分支名称') } stages { stage('📥 代码检出') { steps { script { checkout([ $class: 'GitSCM', branches: [[name: "*/${params.BRANCH}"]], extensions: [], userRemoteConfigs: [[ url: params.REPO_URL, credentialsId: 'github-credentials' ]] ]) } } } stage('🔍 静态分析') { parallel { stage('SonarQube') { steps { sh 'sonar-scanner -Dsonar.login=${SONAR_TOKEN}' } } stage('Semgrep') { steps { sh 'semgrep --config auto --json --output semgrep-results.json' } } stage('CodeQL') { steps { sh 'codeql database create --language=python' sh 'codeql database analyze --format=sarif-latest --output=codeql-results.sarif' } } } } stage('🤖 AI 修复生成') { steps { script { sh """ python -m bugfix_agent.generate \\ --bug-id ${BUG_ID} \\ --repo-path ${WORKSPACE} \\ --output-dir fixes/ \\ --model claude-3-5-sonnet """ } } } stage('🧪 单元测试') { steps { sh 'pytest tests/unit/ --cov=src --cov-report=xml --cov-report=html' junit allowEmptyResults: true, testResults: '**/test-results/*.xml' publishCoverage adapters: [coberturaAdapter('coverage.xml')] } } stage('🔄 回归测试') { steps { sh 'pytest tests/regression/ -v --tb=short' } } stage('🛡️ 安全扫描') { parallel { stage('Snyk') { steps { sh 'snyk test --severity-threshold=high' } } stage('Gitleaks') { steps { sh 'gitleaks detect --report-format json --report-path gitleaks-results.json' } } stage('Bandit') { steps { sh 'bandit -r src/ -f json -o bandit-results.json' } } } } stage('⚡ 性能测试') { steps { sh 'pytest tests/performance/ --benchmark-json=perf-results.json' script { // 对比性能基准 sh 'python compare_benchmarks.py --current perf-results.json --baseline baseline.json' } } } stage('🐳 构建镜像') { steps { script { def buildId = env.BUILD_ID sh "docker build -t app:${buildId} -t app:latest ." } } } stage('🚀 部署到测试环境') { steps { script { sh """ kubectl apply -f k8s/test/ \\ --set image.tag=${BUILD_ID} \\ --namespace test-env """ sh 'kubectl rollout status deployment/app -n test-env --timeout=300s' } } } stage('👁️ E2E 测试') { steps { sh 'npm run e2e -- --env=test' } } stage('📊 生成报告') { steps { script { sh """ python generate_report.py \\ --bug-id ${BUG_ID} \\ --output reports/${BUG_ID}.md \\ --include-logs """ } archiveArtifacts artifacts: 'reports/*.md' } } } post { always { cleanWs() } success { script { // 发送成功通知 sh """ python notify.py \\ --status success \\ --bug-id ${BUG_ID} \\ --channels slack,email,jira """ // 自动合并 PR sh 'gh pr merge --auto --merge' } } failure { script { // 发送失败通知 sh """ python notify.py \\ --status failure \\ --bug-id ${BUG_ID} \\ --channels slack,email """ // 自动回滚 sh 'kubectl rollout undo deployment/app -n test-env' } } unstable { echo '⚠️ 测试不稳定，需要人工审查' } } }

# Prometheus 告警规则 + 自动回滚 groups: - name: bugfix-auto-rollback rules: - alert: HighErrorRate expr: rate(http_requests_total{status=~"5.."}[5m]) > 0.05 for: 2m annotations: summary: "错误率超过 5%，触发自动回滚" rollback: "true" labels: severity: critical - alert: HighLatency expr: histogram_quantile(0.99, rate(http_request_duration_seconds_bucket[5m])) > 2 for: 3m annotations: summary: "P99 延迟超过 2 秒，触发自动回滚" rollback: "true" labels: severity: critical - alert: PodCrashLooping expr: rate(kube_pod_container_status_restarts_total[15m]) > 0 for: 5m annotations: summary: "Pod 频繁重启，触发自动回滚" rollback: "true" labels: severity: critical --- # Argo Rollouts 分析模板 apiVersion: argoproj.io/v1alpha1 kind: AnalysisTemplate metadata: name: error-rate-check spec: metrics: - name: error-rate interval: 1m successCondition: result[0] < 0.05 failureLimit: 1 provider: prometheus: address: http://prometheus:9090 query: | rate(http_requests_total{status=~"5.."}[5m])

ai-bugfix-agent/ ├── README.md ├── requirements.txt ├── Dockerfile ├── docker-compose.yml ├── Makefile │ ├── config/ │ ├── settings.yaml # 系统配置 │ ├── claudemd # Claude Code 规范 │ └── k8s/ │ ├── deployment.yaml │ ├── service.yaml │ ├── configmap.yaml │ └── rbac.yaml │ ├── src/ │ ├── __init__.py │ ├── main.py # 主入口 │ │ │ ├── adapters/ # 多渠道适配器 │ │ ├── __init__.py │ │ ├── github_adapter.py │ │ ├── jira_adapter.py │ │ ├── email_adapter.py │ │ └── slack_adapter.py │ │ │ ├── core/ # 核心引擎 │ │ ├── __init__.py │ │ ├── bug_analyzer.py # Bug 分析器 │ │ ├── fix_generator.py # 修复生成器 │ │ ├── trace_manager.py # Agent Trace 管理 │ │ └── validator.py # 验证引擎 │ │ │ ├── models/ # 数据模型 │ │ ├── __init__.py │ │ ├── bug_event.py │ │ ├── agent_trace.py │ │ └── fix_result.py │ │ │ ├── services/ # 外部服务 │ │ ├── __init__.py │ │ ├── claude_service.py │ │ ├── codex_service.py │ │ ├── git_service.py │ │ └── k8s_service.py │ │ │ └── utils/ # 工具函数 │ ├── __init__.py │ ├── logger.py │ ├── security.py │ └── metrics.py │ ├── tests/ │ ├── unit/ │ ├── integration/ │ ├── regression/ │ └── performance/ │ ├── pipelines/ │ ├── Jenkinsfile │ └── argo-rollouts.yaml │ └── scripts/ ├── setup.sh ├── deploy.sh └── rollback.sh

# src/main.py - 主入口 import asyncio import logging from fastapi import FastAPI, BackgroundTasks from contextlib import asynccontextmanager from .adapters import GitHubAdapter, JiraAdapter, EmailAdapter, SlackAdapter from .core import BugAnalyzer, FixGenerator, TraceManager, Validator from .services import K8sService from .models import BugEvent from .utils import setup_logger, setup_metrics logger = setup_logger(__name__) @asynccontextmanager async def lifespan(app: FastAPI): """应用生命周期管理""" # 启动时初始化 logger.info("🚀 启动 AI Bugfix Agent...") # 初始化服务 app.state.github = GitHubAdapter() app.state.jira = JiraAdapter() app.state.email = EmailAdapter() app.state.slack = SlackAdapter() app.state.analyzer = BugAnalyzer() app.state.generator = FixGenerator() app.state.tracer = TraceManager() app.state.validator = Validator() app.state.k8s = K8sService() yield # 关闭时清理 logger.info("👋 关闭 AI Bugfix Agent...") app = FastAPI( title="AI Bugfix Agent", description="自主 Bug 发现、定位、修复系统", version="2.0.0", lifespan=lifespan ) setup_metrics(app) @app.post("/api/v1/bugs/receive") async def receive_bug( bug_data: dict, background_tasks: BackgroundTasks, source: str = "github" ): """接收 Bug 事件""" logger.info(f"📥 接收 Bug: {source} - {bug_data.get('title')}") # 标准化 Bug 事件 bug_event = BugEvent.from_dict(bug_data, source) # 异步处理 background_tasks.add_task(process_bug, bug_event, app.state) return {"status": "accepted", "bug_id": bug_event.id} async def process_bug(bug_event: BugEvent, state): """处理 Bug 的完整流程""" try: # 1. 去重检查 duplicates = await state.analyzer.check_duplicates(bug_event) if duplicates: logger.info(f"⚠️ 发现重复 Bug: {duplicates}") return # 2. 分析与定位 analysis = await state.analyzer.analyze(bug_event) logger.info(f"🔍 分析完成：{analysis.root_cause}") # 3. 生成修复 fix = await state.generator.generate(bug_event, analysis) logger.info(f"🔧 修复生成完成") # 4. 记录 Agent Trace trace_id = await state.tracer.record(fix) logger.info(f"📝 Trace 记录：{trace_id}") # 5. 验证修复 validation = await state.validator.validate(fix) if not validation.passed: logger.error(f"❌ 验证失败：{validation.errors}") return # 6. 提交修复 pr_url = await state.k8s.submit_fix(fix, bug_event) logger.info(f"✅ 修复提交：{pr_url}") # 7. 通知结果 await notify_success(bug_event, pr_url, validation) except Exception as e: logger.error(f"❌ 处理失败：{e}", exc_info=True) await notify_failure(bug_event, str(e)) if __name__ == "__main__": import uvicorn uvicorn.run(app, host="0.0.0.0", port=8000)

# Dockerfile FROM python:3.11-slim as base # 安全配置 RUN apt-get update && apt-get install -y --no-install-recommends \\ git \\ curl \\ && rm -rf /var/lib/apt/lists/* # 创建非 root 用户 RUN useradd -m -u 1000 -s /bin/bash bugfix-agent WORKDIR /app # 安装依赖 COPY requirements.txt . RUN pip install --no-cache-dir -r requirements.txt # 复制代码 COPY src/ ./src/ COPY config/ ./config/ # 设置权限 RUN chown -R bugfix-agent:bugfix-agent /app USER bugfix-agent # 健康检查 HEALTHCHECK --interval=30s --timeout=10s --start-period=40s --retries=3 \\ CMD curl -f http://localhost:8000/health || exit 1 # 启动命令 CMD ["python", "-m", "uvicorn", "src.main:app", "--host", "0.0.0.0", "--port", "8000"] --- # docker-compose.yml version: '3.8' services: bugfix-agent: build: . ports: - "8000:8000" environment: - ANTHROPIC_API_KEY=${ANTHROPIC_API_KEY} - GITHUB_TOKEN=${GITHUB_TOKEN} - JIRA_API_KEY=${JIRA_API_KEY} volumes: - ./repos:/app/repos:ro - ./logs:/app/logs networks: - bugfix-network deploy: resources: limits: cpus: '2.0' memory: 4G reservations: cpus: '0.5' memory: 1G redis: image: redis:7-alpine networks: - bugfix-network volumes: - redis-data:/data networks: bugfix-network: driver: bridge volumes: redis-data:

# config/k8s/deployment.yaml apiVersion: apps/v1 kind: Deployment metadata: name: ai-bugfix-agent namespace: devops labels: app: bugfix-agent version: v2.0.0 spec: replicas: 3 strategy: type: RollingUpdate rollingUpdate: maxSurge: 1 maxUnavailable: 0 selector: matchLabels: app: bugfix-agent template: metadata: labels: app: bugfix-agent annotations: prometheus.io/scrape: "true" prometheus.io/port: "8000" spec: serviceAccountName: bugfix-agent-sa securityContext: runAsNonRoot: true runAsUser: 1000 fsGroup: 1000 containers: - name: bugfix-agent image: registry.example.com/bugfix-agent:v2.0.0 imagePullPolicy: Always ports: - containerPort: 8000 name: http env: - name: ANTHROPIC_API_KEY valueFrom: secretKeyRef: name: bugfix-secrets key: anthropic-key - name: GITHUB_TOKEN valueFrom: secretKeyRef: name: bugfix-secrets key: github-token resources: requests: cpu: 500m memory: 1Gi limits: cpu: 2000m memory: 4Gi livenessProbe: httpGet: path: /health port: 8000 initialDelaySeconds: 30 periodSeconds: 10 timeoutSeconds: 5 failureThreshold: 3 readinessProbe: httpGet: path: /ready port: 8000 initialDelaySeconds: 10 periodSeconds: 5 timeoutSeconds: 3 failureThreshold: 3 securityContext: allowPrivilegeEscalation: false readOnlyRootFilesystem: true capabilities: drop: - ALL volumeMounts: - name: tmp-volume mountPath: /tmp - name: logs-volume mountPath: /app/logs volumes: - name: tmp-volume emptyDir: {} - name: logs-volume persistentVolumeClaim: claimName: bugfix-logs-pvc affinity: podAntiAffinity: preferredDuringSchedulingIgnoredDuringExecution: - weight: 100 podAffinityTerm: labelSelector: matchLabels: app: bugfix-agent topologyKey: kubernetes.io/hostname --- # config/k8s/service.yaml apiVersion: v1 kind: Service metadata: name: bugfix-agent-service namespace: devops spec: selector: app: bugfix-agent ports: - port: 80 targetPort: 8000 protocol: TCP name: http type: ClusterIP --- # config/k8s/hpa.yaml - 自动扩缩容 apiVersion: autoscaling/v2 kind: HorizontalPodAutoscaler metadata: name: bugfix-agent-hpa namespace: devops spec: scaleTargetRef: apiVersion: apps/v1 kind: Deployment name: ai-bugfix-agent minReplicas: 3 maxReplicas: 10 metrics: - type: Resource resource: name: cpu target: type: Utilization averageUtilization: 70 - type: Resource resource: name: memory target: type: Utilization averageUtilization: 80 behavior: scaleDown: stabilizationWindowSeconds: 300 policies: - type: Percent value: 50 periodSeconds: 60 scaleUp: stabilizationWindowSeconds: 60 policies: - type: Percent value: 100 periodSeconds: 60

# KubeSphere 可视化流水线配置 apiVersion: devops.kubesphere.io/v1alpha3 kind: Pipeline metadata: name: ai-bugfix-pipeline namespace: devops spec: pipelineConfig: type: Jenkinsfile jenkinsfile: | pipeline { agent { kubernetes { yaml ''' apiVersion: v1 kind: Pod spec: containers: - name: bugfix image: bugfix-agent:latest command: ['cat'] tty: true - name: docker image: docker:24.0 command: ['cat'] tty: true ''' } } parameters { string(name: 'BUG_ID', defaultValue: '') string(name: 'REPO_URL', defaultValue: '') } stages { stage('代码检出') { steps { git branch: 'main', url: params.REPO_URL } } stage('AI 修复生成') { steps { sh 'python -m bugfix_agent --bug-id ${BUG_ID}' } } stage('测试验证') { parallel { stage('单元测试') { steps { sh 'pytest tests/unit/ --cov=src' } } stage('安全扫描') { steps { sh 'semgrep --config auto' sh 'snyk test' } } } } stage('构建部署') { steps { sh 'docker build -t app:${BUILD_ID} .' sh 'kubectl apply -f k8s/' } } } } description: "AI Bugfix Agent 自动化修复流水线"