{"id":44321830,"url":"https://github.com/richenlin/evolving-programming-agent","last_synced_at":"2026-02-11T06:35:34.302Z","repository":{"id":334227772,"uuid":"1140480878","full_name":"richenlin/evolving-programming-agent","owner":"richenlin","description":"Evolving Programming Agent - 持续学习进化的编程智能体，由 AI Skills 驱动的自适应编程助手。","archived":false,"fork":false,"pushed_at":"2026-01-31T19:51:33.000Z","size":338,"stargazers_count":2,"open_issues_count":0,"forks_count":0,"subscribers_count":0,"default_branch":"main","last_synced_at":"2026-02-01T07:36:47.867Z","etag":null,"topics":["agent","ai","claude-code","cursor","opencode","skill"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/richenlin.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2026-01-23T10:35:06.000Z","updated_at":"2026-01-31T19:51:37.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/richenlin/evolving-programming-agent","commit_stats":null,"previous_names":["richenlin/evolving-programming-agent"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/richenlin/evolving-programming-agent","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/richenlin%2Fevolving-programming-agent","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/richenlin%2Fevolving-programming-agent/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/richenlin%2Fevolving-programming-agent/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/richenlin%2Fevolving-programming-agent/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/richenlin","download_url":"https://codeload.github.com/richenlin/evolving-programming-agent/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/richenlin%2Fevolving-programming-agent/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":29328270,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-02-11T06:13:03.264Z","status":"ssl_error","status_checked_at":"2026-02-11T06:12:55.843Z","response_time":97,"last_error":"SSL_read: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["agent","ai","claude-code","cursor","opencode","skill"],"created_at":"2026-02-11T06:35:33.459Z","updated_at":"2026-02-11T06:35:34.274Z","avatar_url":"https://github.com/richenlin.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Evolving Programming Agent\n\n**持续学习、自我进化的 AI 编程智能体**\n\nEvolving Programming Agent 是一个模块化的 AI 编程系统。它不仅仅是一个代码生成工具，更是一个能够从 GitHub 学习最佳实践、从日常任务中积累经验、并不断优化自身能力的\"成长型\"智能体。\n\n---\n\n## 🚀 核心特性\n\n- **🧠 统一协调大脑**: 通过 `evolving-agent` 进行意图识别和任务调度，智能处理编程、学习和管理任务\n- **🔄 自动进化闭环**: 在编程任务结束后，自动提取有价值的经验（Bug 修复、架构模式）并存入知识库\n- **📚 GitHub 学习引擎**: 主动分析 GitHub 开源项目，提取架构范式和代码规范，转化为可复用的技能\n- **⚡️ 异步知识流**: 知识检索和归纳在后台异步执行，提供流畅无阻塞的编程体验\n- **🧩 模块化架构**: 所有能力（编程、学习、管理）模块化，职责清晰\n- **🎯 智能触发系统**: 7 大分类的知识存储，基于关键词自动触发相关知识\n- **🌐 跨平台支持**: 支持 OpenCode、Claude Code、Cursor 三大 AI 编程平台\n\n---\n\n## 📋 系统要求\n\n### 必需依赖\n\n| 依赖 | 版本要求 | 说明 |\n|------|----------|------|\n| Python | \u003e= 3.8 | 项目运行环境 |\n| PyYAML | \u003e= 6.0,\u003c 7.0 | 解析 SKILL.md frontmatter |\n\n### 可选依赖\n\n| 依赖 | 版本要求 | 说明 |\n|------|----------|------|\n| Git | \u003e= 2.0 | GitHub 仓库学习功能 |\n\n---\n\n## 🏗️ 架构概览\n\nEvolving Programming Agent 采用**统一协调器模式**（v4.0），通过 `evolving-agent` 作为中央大脑，智能调度各功能模块：\n\n```\n用户输入\n    ↓\nevolving-agent (协调器)\n    ├─► Sequential Thinking 意图识别\n    │   ├─► 编程任务 → programming-assistant\n    │   ├─► GitHub学习 → github-to-skills\n    │   ├─► 知识归纳 → knowledge-base\n    │   └─► 管理命令 → skill-manager\n    ↓\n异步知识检索/归纳\n    ↓\nknowledge-base (统一知识库)\n    ├─► experience (经验)\n    ├─► tech-stack (技术栈)\n    ├─► scenario (场景)\n    ├─► problem (问题)\n    ├─► testing (测试)\n    ├─► pattern (模式)\n    └─► skill (技能)\n```\n\n### 核心组件\n\n| 组件 | 目录 | 职责 |\n|------|------|------|\n| **evolving-agent** | `evolving-agent/` | **核心协调器**。意图识别、任务调度、进化模式管理 |\n| **programming-assistant** | `evolving-agent/modules/programming-assistant/` | 执行引擎。代码生成、修复和重构 |\n| **github-to-skills** | `evolving-agent/modules/github-to-skills/` | 学习引擎。从 GitHub 提取知识 |\n| **knowledge-base** | `evolving-agent/modules/knowledge-base/` | 统一知识库。存储、查询、归纳知识 |\n| **skill-manager** | `skill-manager/` | 运维工具。管理 Skill 的生命周期 |\n\n---\n\n## 📦 安装\n\n### 快速安装\n\n```bash\n# 克隆仓库\ngit clone https://github.com/your-repo/evolving-programming-agent.git\ncd evolving-programming-agent\n\n# 安装所有组件 (支持 OpenCode, Claude Code, Cursor)\n./scripts/install.sh --all\n```\n\n### Python 环境配置\n\n每个 Skill 拥有独立的虚拟环境，避免依赖冲突：\n\n```bash\n# evolving-agent 的虚拟环境\n~/.config/opencode/skills/evolving-agent/.venv/\n\n# skill-manager 复用 evolving-agent 的虚拟环境\n```\n\n---\n\n## 🎮 快速开始\n\n### 1. 启动协调器\n\n推荐使用统一入口命令 `/evolve` 启动会话：\n\n```bash\n/evolve\n```\n\n系统将初始化环境，开启进化模式，并等待您的指令。\n\n### 2. 执行编程任务\n\n直接用自然语言描述需求，协调器会自动调度 `programming-assistant`：\n\n\u003e \"帮我用 React 写一个登录页面\"\n\u003e \"修复这个 CORS 跨域问题\"\n\u003e \"重构这个组件，提高可维护性\"\n\n### 3. 从 GitHub 学习\n\n让智能体学习优秀的开源项目：\n\n\u003e \"学习这个仓库 https://github.com/shadcn/ui\"\n\n系统将自动提取组件设计模式，并在后续编程中复用。\n\n### 4. 显式保存经验\n\n虽然系统会自动进化，您也可以显式要求保存：\n\n\u003e \"记住这个解决方案，以后遇到类似问题直接用\"\n\u003e \"保存这个修复经验，标记为跨域问题\"\n\n---\n\n## 📖 统一命令行接口\n\n所有功能通过 `run.py` 统一调用，支持跨平台（OpenCode/Claude Code/Cursor）：\n\n### 设置路径变量（每个 shell 会话执行一次）\n\n```bash\n# 自动检测平台并设置路径\nSKILLS_DIR=$([ -d ~/.config/opencode/skills/evolving-agent ] \u0026\u0026 echo ~/.config/opencode/skills || echo ~/.claude/skills)\n```\n\n### 进化模式控制\n\n```bash\n# 查看状态\npython $SKILLS_DIR/evolving-agent/scripts/run.py mode --status\n\n# 完整初始化\npython $SKILLS_DIR/evolving-agent/scripts/run.py mode --init\n\n# 开启进化模式\npython $SKILLS_DIR/evolving-agent/scripts/run.py mode --on\n\n# 关闭进化模式\npython $SKILLS_DIR/evolving-agent/scripts/run.py mode --off\n```\n\n### 知识库操作\n\n```bash\n# 查看统计\npython $SKILLS_DIR/evolving-agent/scripts/run.py knowledge query --stats\n\n# 按触发词查询\npython $SKILLS_DIR/evolving-agent/scripts/run.py knowledge query --trigger \"react,hooks\"\n\n# 按分类查询\npython $SKILLS_DIR/evolving-agent/scripts/run.py knowledge query --category problem\n\n# 全文搜索\npython $SKILLS_DIR/evolving-agent/scripts/run.py knowledge query --search \"跨域\"\n\n# 触发检测（自动匹配相关知识）\npython $SKILLS_DIR/evolving-agent/scripts/run.py knowledge trigger --input \"修复CORS问题\"\n\n# 归纳并自动存储\necho \"内容\" | python $SKILLS_DIR/evolving-agent/scripts/run.py knowledge summarize --auto-store\n\n# 存储知识\npython $SKILLS_DIR/evolving-agent/scripts/run.py knowledge store --category experience --name \"xxx\"\n```\n\n### GitHub 学习\n\n```bash\n# 获取仓库信息\npython $SKILLS_DIR/evolving-agent/scripts/run.py github fetch \u003curl\u003e\n\n# 提取模式\npython $SKILLS_DIR/evolving-agent/scripts/run.py github extract --input repo_info.json\n\n# 存储到知识库\npython $SKILLS_DIR/evolving-agent/scripts/run.py github store --input extracted.json\n```\n\n### 项目检测\n\n```bash\n# 检测当前项目技术栈\npython $SKILLS_DIR/evolving-agent/scripts/run.py project detect .\n\n# 存储项目经验\npython $SKILLS_DIR/evolving-agent/scripts/run.py project store --tech react --pattern \"xxx\"\n\n# 查询项目经验\npython $SKILLS_DIR/evolving-agent/scripts/run.py project query --project .\n```\n\n### 环境信息\n\n```bash\n# 显示环境信息（表格格式）\npython $SKILLS_DIR/evolving-agent/scripts/run.py info\n\n# 显示环境信息（JSON 格式）\npython $SKILLS_DIR/evolving-agent/scripts/run.py info --json\n```\n\n---\n\n## 🔄 核心工作流\n\n### 编程 + 进化闭环\n\n```\n用户请求 (\"帮我修复这个Bug\")\n    ↓\nevolving-agent (协调器)\n    │\n    ├─► Sequential Thinking 意图识别\n    ├─► 检查进化模式\n    │\n    ├─► [异步] 知识检索 (knowledge-retrieval)\n    │      ↓\n    │   .knowledge-context.md (上下文)\n    │\n    ▼\nprogramming-assistant (执行器)\n    │\n    ├─► Full Mode / Simple Mode 选择\n    ├─► 任务拆解（sequential-thinking）\n    ├─► 执行修复循环\n    │\n    ▼\n任务完成\n    ↓\nevolving-agent (协调器)\n    │\n    ├─► 检测进化触发条件 (复杂修复? 用户反馈?)\n    │\n    ▼\n[异步] 知识归纳 (knowledge-summarize)\n    ↓\nknowledge-base (存入新经验)\n```\n\n### GitHub 学习闭环\n\n```\n用户请求 (\"学习这个仓库 https://github.com/xx/repo\")\n    ↓\nevolving-agent (协调器)\n    ↓\ngithub-to-skills (学习器)\n    │\n    ├─► Fetch Repo Info\n    ├─► Extract Patterns/Stacks\n    └─► Store to knowledge-base\n          │\n          ▼\nknowledge-base (更新索引)\n    │\n    ▼\n(后续编程任务中自动复用该知识)\n```\n\n---\n\n## 📚 知识分类系统\n\n| 分类 | 目录 | 触发场景 |\n|------|------|----------|\n| **experience** | `experiences/` | 优化、重构、最佳实践 |\n| **tech-stack** | `tech-stacks/` | 框架相关 |\n| **scenario** | `scenarios/` | 创建、实现功能 |\n| **problem** | `problems/` | 修复、调试、报错 |\n| **testing** | `testing/` | 测试相关 |\n| **pattern** | `patterns/` | 架构、设计模式 |\n| **skill** | `skills/` | 通用技巧 |\n\n### 知识条目 Schema\n\n```json\n{\n  \"id\": \"category-name-hash\",\n  \"category\": \"experience|tech-stack|scenario|problem|testing|pattern|skill\",\n  \"name\": \"名称\",\n  \"triggers\": [\"触发词\"],\n  \"content\": {},\n  \"sources\": [\"来源\"],\n  \"created_at\": \"ISO-8601\",\n  \"effectiveness\": 0.5\n}\n```\n\n---\n\n## 🎯 编程模式\n\n### Full Mode（完整开发）\n\n**触发关键词**: 创建、实现、添加、开发\n\n**工作流**:\n1. 任务拆解（sequential-thinking）\n2. 生成 TODO 列表\n3. 执行开发循环\n4. 进化检查\n\n### Simple Mode（快速修复）\n\n**触发关键词**: 修复、fix、bug、重构、优化、review\n\n**工作流**:\n1. 问题分析（sequential-thinking）\n2. 执行修复循环\n3. 进化检查\n\n---\n\n## 📂 项目结构\n\n```\nevolving-programming-agent/\n├── evolving-agent/                 # [Core] 顶层协调器\n│   ├── SKILL.md                    # 协调器配置\n│   ├── command/\n│   │   └── evolve.md              # /evolve 命令文档\n│   ├── scripts/\n│   │   ├── run.py                 # 统一 CLI 入口\n│   │   ├── core/                  # 核心脚本\n│   │   │   ├── toggle_mode.py    # 进化模式控制\n│   │   │   ├── trigger_detector.py # 触发检测\n│   │   │   └── ...\n│   │   ├── knowledge/             # 知识库脚本\n│   │   ├── github/                # GitHub 学习脚本\n│   │   └── programming/           # 编程助手脚本\n│   └── modules/\n│       ├── programming-assistant/ # 执行引擎\n│       │   ├── README.md\n│       │   └── workflows/\n│       │       ├── full-mode.md\n│       │       ├── simple-mode.md\n│       │       └── evolution-check.md\n│       ├── github-to-skills/      # 学习引擎\n│       │   ├── README.md\n│       │   └── ...\n│       └── knowledge-base/        # 统一知识库\n│           ├── README.md\n│           ├── agents/\n│           │   ├── retrieval-agent.md\n│           │   └── summarize-agent.md\n│           └── scripts/\n│\n├── skill-manager/                  # [Util] 运维工具\n│   ├── SKILL.md\n│   └── scripts/\n│       ├── list_skills.py\n│       ├── health_check.py\n│       ├── toggle_skill.py\n│       └── ...\n│\n├── docs/                           # 文档\n│   └── SOLUTION.md                 # 架构设计文档\n├── tests/                          # 测试\n├── scripts/                        # 安装/卸载脚本\n├── requirements.txt                # Python 依赖\n└── README.md                       # 本文件\n```\n\n---\n\n## 🌍 跨平台支持\n\nEvolving Programming Agent 支持三大 AI 编程平台：\n\n| 平台 | Skills 目录 | 知识目录 |\n|------|-------------|----------|\n| **OpenCode** | `~/.config/opencode/skills/` | `~/.config/opencode/knowledge/` |\n| **Claude Code** | `~/.claude/skills/` | `~/.claude/knowledge/` |\n| **Cursor** | `~/.claude/skills/` | `~/.claude/knowledge/` |\n\n系统会自动检测当前平台，并使用正确的路径。\n\n---\n\n## 🤝 贡献\n\n欢迎提交 Pull Request 或 Issue 来帮助改进这个项目！\n\n---\n\n## 📄 许可证\n\nMIT License\n\n---\n\n## 📖 相关文档\n\n- [架构设计 (docs/SOLUTION.md)](docs/SOLUTION.md): 详细的系统架构和设计理念\n- [evolving-agent SKILL](evolving-agent/SKILL.md): 协调器配置文档\n- [skill-manager SKILL](skill-manager/SKILL.md): 运维工具文档\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Frichenlin%2Fevolving-programming-agent","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Frichenlin%2Fevolving-programming-agent","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Frichenlin%2Fevolving-programming-agent/lists"}