Code Quality Analysis Report【免费下载链接】ruflo The original agent meta-harness. Deploy intelligent multi-player swarms, coordinate autonomous workflows, and build conversational AI systems. Features adaptive memory, self-learning intelligence, RAG integration, and native Claude Code / Codex / Hermes and many more Integrated项目地址: https://gitcode.com/GitHub_Trending/cl/rufloSummaryOverall Quality Score: X/10Files Analyzed: NIssues Found: NTechnical Debt Estimate: X hoursCritical Issues[Issue description]File: path/to/file.js:lineSeverity: HighSuggestion: [Improvement]Code SmellsRefactoring OpportunitiesPositive Findings[Good practice observed]模板设计有三个关键点值得注意 1. **Summary 量化**质量分X/10、分析文件数、问题总数、技术债工时估算供上层决策与趋势跟踪 2. **Critical Issues 带定位**每条高危问题必须给出 File: 路径:行号 与 Severity 分级保证可直接派发修复而不仅是模糊描述 3. **Positive Findings 独立成节**避免审查报告全篇负面保留对好实践的肯定这与增强版 communication.include_code_snippets: true、style: technical 的沟通偏好一致。 在 [code-analyzer.md](https://link.gitcode.com/i/47fcf3de528d2d8c181cda3a101a2ba7) 的扩展版本中该模板进一步细化出严重度分级与示例例如对 SQL Injection Risk in UserController.search() 标注 Severity: High 并给出 Use parameterized queries 的修复方向 —— 说明该报告协议可以平滑升级为带修复建议的三级高/中/低缺陷报告。 ## 五、增强版元数据从提示词到可自动触发的声明式 Agent v3 目录下的 [analyze-code-quality.md 增强版](https://link.gitcode.com/i/851c63aaaa1d82edb526daed5e39edb8) 展示了该 Agent 的完整声明式 schema将其从纯提示词升级为可被调度框架自动选路的 Agent。其关键字段如下。 ### 5.1 触发与匹配triggers yaml triggers: keywords: [code review, analyze code, code quality, refactor, technical debt, code smell] file_patterns: [**/*.js, **/*.ts, **/*.py, **/*.java] task_patterns: [review * code, analyze * quality, find code smells] domains: [analysis, quality]keywords当任务文本命中这些词时优先选路file_patterns限定分析目标文件类型task_patterns通配符形式的任务意图匹配domains归属领域便于按领域过滤。5.2 能力与权限边界capabilitiescapabilities: allowed_tools: [Read, Grep, Glob, WebSearch] # WebSearch 仅用于最佳实践调研 restricted_tools: [Write, Edit, MultiEdit, Bash, Task] max_file_operations: 100 max_execution_time: 600 memory_access: both这是该 Agent 的安全设计核心它被设计为只读分析器—— 允许读取/搜索/文件枚举类工具而写文件、执行命令、再委派子任务等全部被禁止。max_file_operations: 100限制单轮操作文件数max_execution_time: 600秒限制最长运行时长memory_access: both表示可读写短期与长期记忆。5.3 路径与文件约束constraintsconstraints: allowed_paths: [src/**, lib/**, app/**, components/**, services/**, utils/**] forbidden_paths: [node_modules/**, .git/**, dist/**, build/**, coverage/**] max_file_size: 1048576 # 1MB allowed_file_types: [.js, .ts, .jsx, .tsx, .py, .java, .go]allowed_paths与forbidden_paths构成了白名单/黑名单双层目录隔离只扫描业务源码目录自动跳过依赖目录、构建产物与 VCS 元数据max_file_size1MB与allowed_file_types则进一步避免分析超大文件或无关二进制。5.4 行为、沟通与集成behavior: error_handling: lenient auto_rollback: false logging_level: verbose communication: style: technical update_frequency: summary include_code_snippets: true emoji_usage: minimal integration: can_delegate_to: [analyze-security, analyze-performance] requires_approval_from: [] shares_context_with: [analyze-refactoring, test-unit]integration.can_delegate_to声明了它的下游协作 Agent安全分析、性能分析用于编排时自动接力shares_context_with声明它与其他 Agent 共享上下文重构分析、单元测试保证分析结论可被无缝交接error_handling: lenient使扫描某个文件失败时不会整体中断配合 on_error 钩子降级为部分分析。5.5 生命周期钩子hookshooks: pre_execution: | echo Code Quality Analyzer initializing... echo Scanning project structure... find . -name *.js -o -name *.ts -o -name *.py | grep -v node_modules | wc -l | xargs echo Files to analyze: echo Checking for code quality configs... ls -la .eslintrc* .prettierrc* .pylintrc tslint.json 2/dev/null || echo No linting configs found post_execution: | echo ✅ Code quality analysis completed echo Analysis stored in memory for future reference echo Run analyze-refactoring for detailed refactoring suggestions on_error: | echo ⚠️ Analysis warning: {{error_message}} echo Continuing with partial analysis...pre_execution启动时清点待分析文件数、探测项目是否已配置.eslintrc*/.prettierrc*/.pylintrc/tslint.json等静态质量工具为后续分析提供环境情报post_execution收尾提示分析完成、结果已写入记忆并引导用户继续运行analyze-refactoring获取深度重构建议 —— 与上面integration.can_delegate_to的接力设计遥相呼应on_error出错时记录警告并继续部分分析契合error_handling: lenient。5.6 触发示例examplesexamples: - trigger: review code quality in the authentication module response: Ill perform a comprehensive code quality analysis of the authentication module, checking for code smells, complexity, and improvement opportunities... - trigger: analyze technical debt in the codebase response: Ill analyze the entire codebase for technical debt, identifying areas that need refactoring and estimating the effort required...这两条示例同时定义了任务意图的解析路径code quality 模块范围 → 代码异味/复杂度/改进点扫描technical debt → 全库技术债盘点与工时估算与回复风格基线可作为单元验证用例使用。六、在 ruflo 多 Agent 体系中的运行与接入code-analyzer 不是孤立文件它被深度嵌入 ruflo 的 Agent 编排框架。6.1 通过 CLI 与 Agent Teams 调度仓库根目录 CLAUDE.md 记录了 ruflo 的多 Agent 执行模式以 Claude Code 的 Task 工具并发 spawning AgentAgent 之间通过命名SendMessage直接通信。CLAUDE.md 给出的典型接力链路为researcher → architect → coder → tester → reviewer其中 reviewer 的职责被描述为 Review code quality and security. Store findings in collaboration.——code-analyzer 正是这一质量审查环的能力提供者。CLAUDE.md 的能力清单中也将code-analyzer列为 init 脚手架预置 Agent 之一与backend-dev、mobile-dev、system-architect、base-template-generator等同列。对应地CLAUDE.md 展示了通用 spawn 语法以 coder 为例npx claude-flowv3alpha agent spawn -t coder --name my-coder【免费下载链接】ruflo The original agent meta-harness. Deploy intelligent multi-player swarms, coordinate autonomous workflows, and build conversational AI systems. Features adaptive memory, self-learning intelligence, RAG integration, and native Claude Code / Codex / Hermes and many more Integrated项目地址: https://gitcode.com/GitHub_Trending/cl/ruflo创作声明：本文部分内容由AI辅助生成（AIGC），仅供参考