app.py

@@ -206,41 +206,51 @@ def handle_pull_request(payload: Dict[str, Any]) -> Tuple[Dict, int]:
                 if web_url:
                     clone_url = web_url.rstrip('/') + '.git'
 
+            # 获取 PR 中变更的文件列表
+            changed_files = []
+            try:
+                if '/' in repo_name:
+                    repo_owner, repo_name_only = repo_name.split('/', 1)
+                else:
+                    repo_owner = 'Bosch_Demo'
+                    repo_name_only = repo_name
+
+                pr_files = gitea_client.get_pull_request_files(repo_owner, repo_name_only, pr_number)
+                if pr_files:
+                    changed_files = [f.get('filename', '') for f in pr_files if f.get('filename')]
+                    logger.info(f"获取到 PR #{pr_number} 的变更文件: {changed_files}")
+            except Exception as e:
+                logger.warning(f"获取 PR 文件列表失败: {e}")
+
             # 执行代码扫描
             scan_results = {}
 
             # Python 扫描
             if 'python' in config.get('scanner', {}).get('languages', []):
                 scan_results['python'] = python_scanner.scan(
-                    clone_url, source_sha, source_branch
+                    clone_url, source_sha, source_branch, changed_files
                 )
 
             # JavaScript/TypeScript 扫描
             if any(lang in config.get('scanner', {}).get('languages', [])
                    for lang in ['javascript', 'typescript']):
                 scan_results['javascript'] = js_scanner.scan(
-                    clone_url, source_sha, source_branch
+                    clone_url, source_sha, source_branch, changed_files
                 )
 
             # 安全扫描
             scan_results['security'] = security_scanner.scan(
-                clone_url, source_sha, source_branch
+                clone_url, source_sha, source_branch, changed_files
             )
 
             # AI 代码审查
             if config.get('ai', {}).get('enabled', False):
                 scan_results['ai'] = ai_reviewer.scan(
-                    clone_url, source_sha, source_branch
+                    clone_url, source_sha, source_branch, changed_files
                 )
 
             # 获取 PR 的代码差异，用于将问题与代码片段关联
             pr_diff = None
-            if '/' in repo_name:
-                repo_owner, repo_name_only = repo_name.split('/', 1)
-            else:
-                repo_owner = 'Bosch_Demo'
-                repo_name_only = repo_name
-
             try:
                 pr_diff = gitea_client.get_pull_request_diff(repo_owner, repo_name_only, pr_number)
                 logger.info(f"已获取 PR #{pr_number} 的 diff，长度: {len(pr_diff) if pr_diff else 0}")
@@ -249,6 +259,12 @@ def handle_pull_request(payload: Dict[str, Any]) -> Tuple[Dict, int]:
 
             # 将问题与代码片段关联
             scan_details_with_code = merge_issues_with_code(scan_results, pr_diff or '')
+            logger.info(f"[DEBUG] scan_results keys: {list(scan_results.keys())}")
+            for k, v in scan_results.items():
+                if isinstance(v, dict):
+                    issues_cnt = len(v.get('issues', []))
+                    logger.info(f"[DEBUG] scan_results['{k}'] issues count: {issues_cnt}")
+            logger.info(f"[DEBUG] scan_details_with_code scanners: {[s.get('name') for s in scan_details_with_code.get('scanners', [])] if scan_details_with_code else 'None'}")
 
             # 生成报告
             commit_message = f'PR #{pr_number}: {pr_title}'
@@ -580,12 +596,24 @@ def api_get_pr_file_content(pr_id):
         # 获取该文件的扫描问题（PR 创建时已扫描并存入 scan_details_with_code）
         scan_issues = []
         path_norm = path.replace('\\', '/').strip()
+        logger.info(f"[DEBUG] 请求文件: path_norm={path_norm}")
         scan_details = pr.get('scan_details_with_code')
         if isinstance(scan_details, str):
             try:
                 scan_details = json.loads(scan_details)
             except Exception:
                 scan_details = None
+        if scan_details:
+            logger.info(f"[DEBUG] scan_details keys: {list(scan_details.keys()) if isinstance(scan_details, dict) else 'not dict'}")
+            if scan_details.get('scanners'):
+                logger.info(f"[DEBUG] scanners count: {len(scan_details['scanners'])}")
+                for scanner in scan_details['scanners']:
+                    scanner_name = scanner.get('name', '')
+                    issues_count = len(scanner.get('issues', []))
+                    logger.info(f"[DEBUG] scanner={scanner_name}, issues_count={issues_count}")
+                    # 打印前几个 issue 的 file 看看
+                    for idx, issue in enumerate(scanner.get('issues', [])[:3]):
+                        logger.info(f"[DEBUG]   issue[{idx}] file={issue.get('file')}, line={issue.get('line')}")
         if scan_details and scan_details.get('scanners'):
             for scanner in scan_details['scanners']:
                 for issue in scanner.get('issues', []):
@@ -594,6 +622,7 @@ def api_get_pr_file_content(pr_id):
                         continue
                     # 匹配：精确相等或一端包含另一端（兼容 basename 或完整路径）
                     if path_norm == issue_file or path_norm.endswith(issue_file) or issue_file.endswith(path_norm):
+                        logger.info(f"[DEBUG] 匹配成功: issue_file={issue_file}, path_norm={path_norm}")
                         sev = (issue.get('severity') or 'info')
                         if isinstance(sev, str):
                             sev = sev.lower()
@@ -606,8 +635,33 @@ def api_get_pr_file_content(pr_id):
                             'message': (issue.get('message') or issue.get('description') or '').strip(),
                             'code_context': issue.get('code_context')
                         })
+        logger.info(f"[DEBUG] 最终 scan_issues count: {len(scan_issues)}")
 
-        return jsonify({'path': path, 'content': content, 'scan_issues': scan_issues})
+        # 获取 AI 审查结果
+        ai_issues = []
+        if scan_details and scan_details.get('ai'):
+            ai_data = scan_details['ai']
+            for issue in ai_data.get('issues', []):
+                issue_file = (issue.get('file') or '').replace('\\', '/').strip()
+                if not issue_file:
+                    continue
+                # 匹配：精确相等或一端包含另一端
+                if path_norm == issue_file or path_norm.endswith(issue_file) or issue_file.endswith(path_norm):
+                    ai_issues.append({
+                        'scanner': 'AI',
+                        'severity': issue.get('severity', 'info'),
+                        'line': int(issue.get('line') or 1),
+                        'message': issue.get('message', ''),
+                        'category': 'ai',
+                        'code_context': issue.get('code_context')
+                    })
+
+        logger.info(f"[DEBUG] AI issues count: {len(ai_issues)}")
+
+        # 合并静态扫描问题和 AI 问题
+        all_issues = scan_issues + ai_issues
+
+        return jsonify({'path': path, 'content': content, 'scan_issues': all_issues})
     except Exception as e:
         logger.error(f'获取文件内容失败: {str(e)}')
         return jsonify({'error': str(e)}), 500

config.yaml

@@ -49,8 +49,8 @@ ai:
   # AI 审查器配置
   # 支持: "ollama" (本地) 或 "api" (在线API)
   provider: "api"
-  # 模型名称（硅基流动可用模型）
-  model: "Qwen/Qwen2.5-7B-Instruct"
+  # 模型名称（硅基流动可用模型）- Qwen 最强语言模型
+  model: "deepseek-ai/DeepSeek-V3.2"
   # API 地址
   # 硅基流动: https://api.siliconflow.cn/v1
   api_url: "https://api.siliconflow.cn/v1"

scanner/ai_reviewer.py

@@ -41,7 +41,7 @@ class AIReviewer(BaseScanner):
 
         logger.info(f'AI 审查器初始化: {self.provider}/{self.model}')
 
-    def scan(self, repo_url: str, commit_id: Optional[str], branch: str) -> Dict[str, Any]:
+    def scan(self, repo_url: str, commit_id: Optional[str], branch: str, changed_files: Optional[List[str]] = None) -> Dict[str, Any]:
         """
         执行代码扫描（实现抽象方法）
 
@@ -49,16 +49,18 @@ class AIReviewer(BaseScanner):
             repo_url: 仓库 URL
             commit_id: 提交 ID
             branch: 分支名
+            changed_files: 可选的变更文件列表（来自 PR）
 
         Returns:
             审查结果
         """
         # 调用实际的审查逻辑
-        return self._do_review(repo_url=repo_url, commit_id=commit_id, branch=branch)
+        return self._do_review(repo_url=repo_url, commit_id=commit_id, branch=branch, changed_files=changed_files)
 
     def _do_review(self, clone_dir: str = None, repo_url: str = None,
                    commit_id: str = None, branch: str = None,
-                   language: str = 'python') -> Dict[str, Any]:
+                   language: str = 'python',
+                   changed_files: Optional[List[str]] = None) -> Dict[str, Any]:
         """
         执行 AI 代码审查
 
@@ -68,6 +70,7 @@ class AIReviewer(BaseScanner):
             commit_id: 提交 ID
             branch: 分支名
             language: 编程语言
+            changed_files: 可选的变更文件列表（来自 PR）
 
         Returns:
             审查结果
@@ -94,7 +97,7 @@ class AIReviewer(BaseScanner):
                 }
 
             # 获取要审查的代码文件
-            files = self._get_code_files(clone_dir, language)
+            files = self._get_code_files(clone_dir, language, changed_files)
 
             if not files:
                 return {
@@ -107,7 +110,7 @@ class AIReviewer(BaseScanner):
             # 对每个文件进行 AI 审查
             all_reviews = []
             for file_path in files[:5]:  # 限制最多审查 5 个文件
-                review = self._review_file(file_path, language)
+                review = self._review_file(file_path, language, clone_dir)
                 if review:
                     all_reviews.append(review)
 
@@ -133,7 +136,7 @@ class AIReviewer(BaseScanner):
                 'summary': f'AI 审查出错: {str(e)}'
             }
 
-    def _get_code_files(self, clone_dir: str, language: str) -> List[str]:
+    def _get_code_files(self, clone_dir: str, language: str, changed_files: Optional[List[str]] = None) -> List[str]:
         """获取代码文件列表"""
         import glob
 
@@ -144,6 +147,18 @@ class AIReviewer(BaseScanner):
         }
 
         exts = extensions.get(language, ['.py'])
+
+        # 如果提供了变更文件列表，只返回这些文件
+        if changed_files:
+            files = []
+            for changed_file in changed_files:
+                if any(changed_file.endswith(ext) for ext in exts):
+                    full_path = os.path.join(clone_dir, changed_file)
+                    if os.path.exists(full_path):
+                        files.append(full_path)
+            return files[:10]
+
+        # 否则扫描整个仓库
         files = []
 
         for ext in exts:
@@ -157,7 +172,7 @@ class AIReviewer(BaseScanner):
 
         return files[:10]  # 最多 10 个文件
 
-    def _review_file(self, file_path: str, language: str) -> Optional[Dict[str, Any]]:
+    def _review_file(self, file_path: str, language: str, clone_dir: str = None) -> Optional[Dict[str, Any]]:
         """审查单个文件"""
         try:
             with open(file_path, 'r', encoding='utf-8') as f:
@@ -181,9 +196,9 @@ class AIReviewer(BaseScanner):
                 return None
 
             # 解析响应
-            filename = os.path.basename(file_path)
+            rel_path = os.path.relpath(file_path, clone_dir) if (clone_dir and file_path) else file_path
             return {
-                'file': filename,
+                'file': rel_path,
                 'path': file_path,
                 'truncated': truncated,
                 'review': response
@@ -236,6 +251,7 @@ class AIReviewer(BaseScanner):
                 logger.warning(f'未知的 AI provider: {self.provider}')
                 return None
         except Exception as e:
+            print("异常追踪信息:", e.__traceback__)
             logger.error(f'AI 调用失败: {str(e)}')
             return None
 

scanner/base.py

@@ -152,15 +152,28 @@ class BaseScanner(ABC):
                 'stdout': '',
                 'stderr': str(e)
             }
-    def get_changed_files(self, clone_dir: str, extensions: List[str]) -> List[str]:
+    def get_changed_files(self, clone_dir: str, extensions: List[str], changed_files: Optional[List[str]] = None) -> List[str]:
         """
         获取指定扩展名的文件列表
         Args:
             clone_dir: 仓库目录
             extensions: 文件扩展名列表
+            changed_files: 可选的变更文件列表（来自 PR），如果提供则只返回这些文件
         Returns:
             文件路径列表
         """
+        # 如果提供了变更文件列表，只扫描这些文件
+        if changed_files:
+            files = []
+            for changed_file in changed_files:
+                # 检查文件扩展名是否匹配
+                if any(changed_file.endswith(ext) for ext in extensions):
+                    full_path = os.path.join(clone_dir, changed_file)
+                    if os.path.exists(full_path):
+                        files.append(full_path)
+            return files
+
+        # 否则扫描整个仓库
         files = []
         for root, dirs, filenames in os.walk(clone_dir):
             # 跳过隐藏目录和特殊目录

scanner/diff_parser.py

@@ -94,10 +94,8 @@ class DiffParser:
 
 def merge_issues_with_code(scan_results: Dict[str, Any], diff: str) -> Dict[str, Any]:
     """将扫描问题与代码片段关联"""
-    if not diff:
-        return scan_results
+    parser = DiffParser(diff) if diff else None
 
-    parser = DiffParser(diff)
     enriched_results = {
         'scanners': [],
         'summary': scan_results.get('summary', {}),
@@ -118,17 +116,93 @@ def merge_issues_with_code(scan_results: Dict[str, Any], diff: str) -> Dict[str,
 
             issues = scanner_data.get('issues', [])
             for issue in issues:
-                enriched_issue = enrich_issue_with_code(issue, parser)
+                enriched_issue = enrich_issue_with_code(issue, parser) if parser else issue
                 enriched_scanner['issues'].append(enriched_issue)
 
             enriched_results['scanners'].append(enriched_scanner)
 
+    # 处理 AI 审查结果，转换为问题格式
     if 'ai' in scan_results:
-        enriched_results['ai'] = scan_results['ai']
+        ai_issues = convert_ai_reviews_to_issues(scan_results['ai'], parser)
+        enriched_results['ai'] = {
+            'name': 'ai',
+            'issues': ai_issues,
+            'summary': scan_results['ai'].get('summary', ''),
+            'files_reviewed': scan_results['ai'].get('files_reviewed', 0)
+        }
 
     return enriched_results
 
 
+def convert_ai_reviews_to_issues(ai_result: Dict[str, Any], parser: Optional[DiffParser] = None) -> List[Dict[str, Any]]:
+    """将 AI 审查结果转换为问题格式"""
+    issues = []
+
+    reviews = ai_result.get('reviews', [])
+    for review in reviews:
+        file_path = review.get('file', '')
+        review_data = review.get('review', {})
+
+        if not review_data:
+            continue
+
+        # 获取文件内容作为代码上下文
+        code_context = None
+        if parser:
+            matched_path = None
+            for path in parser.files.keys():
+                if file_path.endswith(path) or path.endswith(file_path) or file_path in path:
+                    matched_path = path
+                    break
+
+            if matched_path:
+                chunk = parser.get_file_content(matched_path)
+                if chunk and chunk.new_content:
+                    lines = chunk.new_content.split('\n')[:10]
+                    code_context = {
+                        'file': matched_path,
+                        'line': 1,
+                        'preview': '\n'.join(lines),
+                        'has_more': len(chunk.new_content.split('\n')) > 10
+                    }
+
+        # 处理优点（不作为问题显示）
+        advantages = review_data.get('优点', [])
+        # 处理问题
+        problems = review_data.get('问题', [])
+        for idx, problem in enumerate(problems):
+            issues.append({
+                'file': file_path,
+                'line': 1,  # AI 审查不返回具体行号
+                'severity': 'warning',
+                'message': f'[AI 建议] {problem}',
+                'category': 'ai',
+                'code_context': code_context,
+                'review_data': {
+                    'type': '问题',
+                    'content': problem
+                }
+            })
+
+        # 处理优化建议
+        optimizations = review_data.get('优化', [])
+        for optimization in optimizations:
+            issues.append({
+                'file': file_path,
+                'line': 1,
+                'severity': 'info',
+                'message': f'[AI 优化] {optimization}',
+                'category': 'ai',
+                'code_context': code_context,
+                'review_data': {
+                    'type': '优化',
+                    'content': optimization
+                }
+            })
+
+    return issues
+
+
 def enrich_issue_with_code(issue: Dict[str, Any], parser: DiffParser) -> Dict[str, Any]:
     """为单个问题添加代码片段"""
     enriched = issue.copy()

scanner/js_scanner.py

@@ -20,7 +20,7 @@ class JavaScriptScanner(BaseScanner):
         super().__init__(config)
         self.extensions = ['.js', '.jsx', '.ts', '.tsx', '.vue', '.svelte']
 
-    def scan(self, repo_url: str, commit_id: Optional[str], branch: str) -> Dict[str, Any]:
+    def scan(self, repo_url: str, commit_id: Optional[str], branch: str, changed_files: Optional[List[str]] = None) -> Dict[str, Any]:
         """
         执行 JavaScript/TypeScript 代码扫描
 
@@ -28,6 +28,7 @@ class JavaScriptScanner(BaseScanner):
             repo_url: 仓库 URL
             commit_id: 提交 ID
             branch: 分支名
+            changed_files: 可选的变更文件列表（来自 PR）
 
         Returns:
             扫描结果
@@ -51,8 +52,8 @@ class JavaScriptScanner(BaseScanner):
             # 克隆仓库
             clone_dir = self.clone_repo(repo_url, commit_id, branch)
 
-            # 获取 JavaScript/TypeScript 文件
-            js_files = self.get_changed_files(clone_dir, self.extensions)
+            # 获取 JavaScript/TypeScript 文件（只扫描变更的文件）
+            js_files = self.get_changed_files(clone_dir, self.extensions, changed_files)
             result['files_scanned'] = len(js_files)
 
             if not js_files:
@@ -75,7 +76,7 @@ class JavaScriptScanner(BaseScanner):
 
         return result
 
-    def _run_eslint(self, cwd: str, files: List[str]) -> Dict[str, Any]:
+    def _run_eslint(self, clone_dir: str, files: List[str]) -> Dict[str, Any]:
         """运行 ESLint 扫描"""
         result = {
             'tool': 'eslint',
@@ -88,7 +89,7 @@ class JavaScriptScanner(BaseScanner):
             cmd = ['npx', 'eslint', '--format=json', '--no-eslintrc'] + files
             
             # 如果没有 eslint 配置，先创建默认配置
-            eslintrc_path = os.path.join(cwd, '.eslintrc.json')
+            eslintrc_path = os.path.join(clone_dir, '.eslintrc.json')
             if not os.path.exists(eslintrc_path):
                 # 创建简单的 ESLint 配置
                 eslint_config = {
@@ -106,7 +107,7 @@ class JavaScriptScanner(BaseScanner):
                 with open(eslintrc_path, 'w') as f:
                     json.dump(eslint_config, f)
 
-            output = self.run_command(cmd, cwd, timeout=120)
+            output = self.run_command(cmd, clone_dir, timeout=120)
             result['raw_output'] = output.get('stdout', '') + output.get('stderr', '')
 
             # 解析 JSON 输出
@@ -115,6 +116,8 @@ class JavaScriptScanner(BaseScanner):
                     eslint_results = json.loads(output['stdout'])
                     for file_result in eslint_results:
                         file_path = file_result.get('filePath', '')
+                        # 使用相对于 clone_dir 的路径
+                        rel_path = os.path.relpath(file_path, clone_dir) if file_path else ''
                         messages = file_result.get('messages', [])
                         
                         for msg in messages:
@@ -124,7 +127,7 @@ class JavaScriptScanner(BaseScanner):
                                 'type': severity,
                                 'severity': 'Error' if msg.get('severity', 0) == 2 else 'Warning',
                                 'message': msg.get('message', ''),
-                                'file': os.path.basename(file_path),
+                                'file': rel_path,
                                 'line': msg.get('line', 0),
                                 'column': msg.get('column', 0),
                                 'symbol': msg.get('ruleId', 'unknown')

scanner/python_scanner.py

@@ -20,7 +20,7 @@ class PythonScanner(BaseScanner):
         super().__init__(config)
         self.extensions = ['.py']
 
-    def scan(self, repo_url: str, commit_id: Optional[str], branch: str) -> Dict[str, Any]:
+    def scan(self, repo_url: str, commit_id: Optional[str], branch: str, changed_files: Optional[List[str]] = None) -> Dict[str, Any]:
         """
         执行 Python 代码扫描
 
@@ -28,6 +28,7 @@ class PythonScanner(BaseScanner):
             repo_url: 仓库 URL
             commit_id: 提交 ID
             branch: 分支名
+            changed_files: 可选的变更文件列表（来自 PR）
 
         Returns:
             扫描结果
@@ -51,8 +52,8 @@ class PythonScanner(BaseScanner):
             # 克隆仓库
             clone_dir = self.clone_repo(repo_url, commit_id, branch)
 
-            # 获取 Python 文件
-            py_files = self.get_changed_files(clone_dir, self.extensions)
+            # 获取 Python 文件（只扫描变更的文件）
+            py_files = self.get_changed_files(clone_dir, self.extensions, changed_files)
             result['files_scanned'] = len(py_files)
 
             if not py_files:
@@ -84,7 +85,7 @@ class PythonScanner(BaseScanner):
 
         return result
 
-    def _run_pylint(self, cwd: str, files: List[str]) -> Dict[str, Any]:
+    def _run_pylint(self, clone_dir: str, files: List[str]) -> Dict[str, Any]:
         """运行 Pylint 扫描"""
         result = {
             'tool': 'pylint',
@@ -95,7 +96,7 @@ class PythonScanner(BaseScanner):
         # 只扫描变更的文件
         try:
             cmd = ['python', '-m', 'pylint', '--output-format=json'] + files
-            output = self.run_command(cmd, cwd, timeout=120)
+            output = self.run_command(cmd, clone_dir, timeout=120)
 
             result['raw_output'] = output.get('stdout', '')
 
@@ -104,12 +105,15 @@ class PythonScanner(BaseScanner):
                 try:
                     issues = json.loads(output['stdout'])
                     for issue in issues:
+                        # 使用相对于 clone_dir 的路径
+                        full_path = issue.get('path', '')
+                        rel_path = os.path.relpath(full_path, clone_dir) if full_path else ''
                         result['issues'].append({
                             'tool': 'pylint',
                             'type': issue.get('type', 'info'),
                             'severity': issue.get('severity', 'Info'),
                             'message': issue.get('message', ''),
-                            'file': os.path.basename(issue.get('path', '')),
+                            'file': rel_path,
                             'line': issue.get('line', 0),
                             'column': issue.get('column', 0),
                             'symbol': issue.get('symbol', '')
@@ -122,7 +126,7 @@ class PythonScanner(BaseScanner):
 
         return result
 
-    def _run_flake8(self, cwd: str, files: List[str]) -> Dict[str, Any]:
+    def _run_flake8(self, clone_dir: str, files: List[str]) -> Dict[str, Any]:
         """运行 Flake8 扫描"""
         result = {
             'tool': 'flake8',
@@ -132,7 +136,7 @@ class PythonScanner(BaseScanner):
 
         try:
             cmd = ['python', '-m', 'flake8', '--format=json'] + files
-            output = self.run_command(cmd, cwd, timeout=120)
+            output = self.run_command(cmd, clone_dir, timeout=120)
 
             result['raw_output'] = output.get('stdout', '')
 
@@ -141,12 +145,15 @@ class PythonScanner(BaseScanner):
                 try:
                     issues = json.loads(output['stdout'])
                     for issue in issues:
+                        # 使用相对于 clone_dir 的路径
+                        full_path = issue.get('filename', '')
+                        rel_path = os.path.relpath(full_path, clone_dir) if full_path else ''
                         result['issues'].append({
                             'tool': 'flake8',
                             'type': self._map_flake8_code(issue.get('code', '')),
                             'severity': 'Warning',
                             'message': issue.get('text', ''),
-                            'file': os.path.basename(issue.get('filename', '')),
+                            'file': rel_path,
                             'line': issue.get('line_number', 0),
                             'column': issue.get('column_number', 0),
                             'symbol': issue.get('code', '')

scanner/security_scanner.py

@@ -21,7 +21,7 @@ class SecurityScanner(BaseScanner):
         # 扫描所有代码文件以发现安全问题
         self.extensions = ['.py', '.js', '.ts', '.jsx', '.tsx', '.java', '.go', '.rb', '.php']
 
-    def scan(self, repo_url: str, commit_id: Optional[str], branch: str) -> Dict[str, Any]:
+    def scan(self, repo_url: str, commit_id: Optional[str], branch: str, changed_files: Optional[List[str]] = None) -> Dict[str, Any]:
         """
         执行安全扫描
 
@@ -29,6 +29,7 @@ class SecurityScanner(BaseScanner):
             repo_url: 仓库 URL
             commit_id: 提交 ID
             branch: 分支名
+            changed_files: 可选的变更文件列表（来自 PR）
 
         Returns:
             扫描结果
@@ -53,8 +54,8 @@ class SecurityScanner(BaseScanner):
             # 克隆仓库
             clone_dir = self.clone_repo(repo_url, commit_id, branch)
 
-            # 获取所有支持的文件
-            all_files = self.get_changed_files(clone_dir, self.extensions)
+            # 获取所有支持的文件（只扫描变更的文件）
+            all_files = self.get_changed_files(clone_dir, self.extensions, changed_files)
             result['files_scanned'] = len(all_files)
 
             if not all_files:
@@ -88,7 +89,7 @@ class SecurityScanner(BaseScanner):
 
         return result
 
-    def _run_bandit(self, cwd: str, files: List[str]) -> Dict[str, Any]:
+    def _run_bandit(self, clone_dir: str, files: List[str]) -> Dict[str, Any]:
         """运行 Bandit 安全扫描"""
         result = {
             'tool': 'bandit',
@@ -98,7 +99,7 @@ class SecurityScanner(BaseScanner):
         try:
             # 运行 bandit
             cmd = ['python', '-m', 'bandit', '-f', 'json'] + files
-            output = self.run_command(cmd, cwd, timeout=120)
+            output = self.run_command(cmd, clone_dir, timeout=120)
 
             # 解析 JSON 输出
             if output.get('stdout'):
@@ -107,6 +108,9 @@ class SecurityScanner(BaseScanner):
                     results = data.get('results', [])
                     
                     for issue in results:
+                        # 使用相对于 clone_dir 的路径
+                        full_path = issue.get('filename', '')
+                        rel_path = os.path.relpath(full_path, clone_dir) if full_path else ''
                         # 映射严重级别
                         severity = issue.get('issue_severity', 'LOW')
                         result['issues'].append({
@@ -115,7 +119,7 @@ class SecurityScanner(BaseScanner):
                             'severity': severity,
                             'confidence': issue.get('issue_confidence', 'LOW'),
                             'message': issue.get('issue_text', ''),
-                            'file': os.path.basename(issue.get('filename', '')),
+                            'file': rel_path,
                             'line': issue.get('line_number', 0),
                             'code': issue.get('code', '')
                         })
@@ -127,7 +131,7 @@ class SecurityScanner(BaseScanner):
 
         return result
 
-    def _scan_js_security(self, cwd: str, files: List[str]) -> Dict[str, Any]:
+    def _scan_js_security(self, clone_dir: str, files: List[str]) -> Dict[str, Any]:
         """简单的 JavaScript 安全扫描（基于模式匹配）"""
         result = {
             'tool': 'js-security',
@@ -172,6 +176,8 @@ class SecurityScanner(BaseScanner):
 
         for file_path in files:
             try:
+                # 使用相对于 clone_dir 的路径
+                rel_path = os.path.relpath(file_path, clone_dir) if file_path else ''
                 with open(file_path, 'r', encoding='utf-8', errors='ignore') as f:
                     content = f.read()
                     lines = content.split('\n')
@@ -185,7 +191,7 @@ class SecurityScanner(BaseScanner):
                                     'severity': pattern_info['severity'],
                                     'confidence': 'MEDIUM',
                                     'message': pattern_info['message'],
-                                    'file': os.path.basename(file_path),
+                                    'file': rel_path,
                                     'line': line_num,
                                     'code': line.strip()[:80]
                                 })