{"id":50395365,"url":"https://github.com/sjtu-sai-agents/EvoMaster","last_synced_at":"2026-06-16T12:00:32.230Z","repository":{"id":335141187,"uuid":"1142570582","full_name":"sjtu-sai-agents/EvoMaster","owner":"sjtu-sai-agents","description":"EvoMaster: A Foundational Evolving Agent Framework for Agentic Science at Scale.","archived":false,"fork":false,"pushed_at":"2026-05-18T13:13:30.000Z","size":16461,"stargazers_count":176,"open_issues_count":3,"forks_count":33,"subscribers_count":3,"default_branch":"main","last_synced_at":"2026-05-18T15:18:35.065Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"apache-2.0","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/sjtu-sai-agents.png","metadata":{"files":{"readme":"README-zh.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","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":"2026-01-26T15:28:07.000Z","updated_at":"2026-05-18T14:45:34.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/sjtu-sai-agents/EvoMaster","commit_stats":null,"previous_names":["sjtu-sai-agents/evomaster"],"tags_count":2,"template":false,"template_full_name":null,"purl":"pkg:github/sjtu-sai-agents/EvoMaster","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/sjtu-sai-agents%2FEvoMaster","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/sjtu-sai-agents%2FEvoMaster/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/sjtu-sai-agents%2FEvoMaster/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/sjtu-sai-agents%2FEvoMaster/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/sjtu-sai-agents","download_url":"https://codeload.github.com/sjtu-sai-agents/EvoMaster/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/sjtu-sai-agents%2FEvoMaster/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":34404748,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-05-26T15:22:16.424Z","status":"online","status_checked_at":"2026-06-16T02:00:06.860Z","response_time":126,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"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":[],"created_at":"2026-05-30T21:00:21.863Z","updated_at":"2026-06-16T12:00:32.224Z","avatar_url":"https://github.com/sjtu-sai-agents.png","language":"Python","funding_links":[],"categories":["Frameworks \u0026 Platforms"],"sub_categories":[],"readme":"\u003cp align=\"center\"\u003e\n  \u003cimg src=\"./assets/LOGO.png\" alt=\"EvoMaster Logo\" width=\"\" /\u003e\n\u003c/p\u003e\n\n\u003cp align=\"center\"\u003e\n  【\u003ca href=\"./README.md\"\u003eEnglish\u003c/a\u003e | \u003ca href=\"./README-zh.md\"\u003e简体中文\u003c/a\u003e】\n\u003c/p\u003e\n\u003cp align=\"center\"\u003e\n  \u003ca href=\"#quick-start\"\u003e\u003cimg src=\"https://img.shields.io/badge/快速开始-3分钟上手-0ea5e9?style=for-the-badge\" alt=\"快速开始\"\u003e\u003c/a\u003e\n  \u003ca href=\"#scimaster-series\"\u003e\u003cimg src=\"https://img.shields.io/badge/SciMaster-6%2B智能体-059669?style=for-the-badge\" alt=\"SciMaster\"\u003e\u003c/a\u003e\n  \u003ca href=\"#key-features\"\u003e\u003cimg src=\"https://img.shields.io/badge/核心特性-5项-7c3aed?style=for-the-badge\" alt=\"核心特性\"\u003e\u003c/a\u003e\n  \u003ca href=\"LICENSE\"\u003e\u003cimg src=\"https://img.shields.io/badge/License-Apache%202.0-ea580c?style=for-the-badge\" alt=\"License\"\u003e\u003c/a\u003e\n  \u003ca href=\"https://arxiv.org/abs/2604.17406\"\u003e\u003cimg src=\"https://img.shields.io/badge/arXiv-2604.17406-b31b1b?style=for-the-badge\u0026logo=arxiv\u0026logoColor=white\" alt=\"arXiv\"\u003e\u003c/a\u003e\n  \u003ca href=\"#contact-us\"\u003e\u003cimg src=\"https://img.shields.io/badge/微信-加入社群-07C160?style=for-the-badge\u0026logo=wechat\u0026logoColor=white\" alt=\"WeChat Group\"\u003e\u003c/a\u003e\n\u003c/p\u003e\n\u003cdiv align=\"center\"\u003e\n\n**面向规模化自主科研（Agentic Science at Scale）的通用自动科研智能体基座**\n\n*让科学智能体开发更简单、模块化且功能强大，加速\"AI for Science\"的变革进程。*\n\n\u003cp align=\"center\"\u003e\n  \u003cimg src=\"./assets/evomaster_vs_openclaw.png\" alt=\"EvoMaster 与 OpenClaw 在四个基准上的性能对比\" width=\"88%\" /\u003e\u003cbr\u003e\n  \u003cem\u003eEvoMaster 在四个权威基准上均优于通用智能体 OpenClaw（在相同 GPT-5.4 后端下）。\u003c/em\u003e\n\u003c/p\u003e\n\n\u003cp align=\"center\"\u003e\n  \u003cimg src=\"./assets/EvoMaster_FSscore.png\" alt=\"EvoMaster 在各基准上的 FS 分数\" width=\"88%\" /\u003e\u003cbr\u003e\n  \u003cem\u003eEvoMaster 在 OpenAI Frontier Science Benchmark 上取得突破。\u003c/em\u003e\n\u003c/p\u003e\n\n\n*基于EvoMaster实现的科研工作流的完整闭环*\n\n\u003ctable align=\"center\" width=\"100%\"\u003e\n\u003ctr\u003e\n\u003ctd width=\"33%\" align=\"center\" style=\"vertical-align: top; padding: 10px;\"\u003e\n\n\u003cstrong\u003e大模型训练\u003c/strong\u003e\n\nhttps://github.com/user-attachments/assets/a4b8042a-72ab-4fc3-893b-243e74eca6cc\n\n\u003c/td\u003e\n\u003ctd width=\"33%\" align=\"center\" style=\"vertical-align: top; padding: 10px;\"\u003e\n\n\u003cstrong\u003e材料科学 (Material Science)\u003c/strong\u003e\n\nhttps://github.com/user-attachments/assets/43b94144-a684-4ffe-9c45-aa5fd2242ce2\n\n\u003c/td\u003e\n\u003ctd width=\"33%\" align=\"center\" style=\"vertical-align: top; padding: 10px;\"\u003e\n\n\u003cstrong\u003e创建 ML 智能体 (Create an ML Agent)\u003c/strong\u003e\n\nhttps://github.com/user-attachments/assets/a8fe00ba-531c-4d53-b7bd-1bbedf7a6442\n\n\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003c/table\u003e\n\n\u003c/div\u003e\n\n## 📰 News\n\n**2026-05-18** EvoMaster 支持运行级自进化功能。智能体可以分析已完成的轨迹，生成 skills, tools和prompt overlays，记录详细 evolution 日志，并自动用进化后的 overlay 再次运行。详情请见[自进化使用指南](./docs/zh/evolution.md)。\n\n**2026-04-19** EvoMaster 预印本版本正式发布！欢迎查看我们的 [arXiv 论文](https://arxiv.org/abs/2604.17406)。\n\n**2026-04-12** EvoMaster `v0.1.1` 发布！MagiClaw现已移至[独立仓库](https://github.com/sjtu-sai-agents/MagiClaw)。EvoMaster 现已支持以skill形式被调用！添加示例自定义工具。\n\n**2026-03-23** EvoMaster `v0.1.0` 发布！开源ML-Master 1.0，ML-Master 2.0，X-Master，Browse-Master 等智能体，支持[MagiClaw](https://github.com/sjtu-sai-agents/MagiClaw)：以飞书对话形式创建智能体，支持typescript形式skill。\n\n**2026-03-02** EvoMaster `v0.0.2` 发布！代码进行重构优化，智能体支持高度自定义。\n\n**2026-02-06** EvoMaster 初版代码 `v0.0.1` 发布！\n\n## \u003ca id=\"introduction\"\u003e\u003c/a\u003e📖 项目介绍\n\n**EvoMaster** 是一个轻量级但功能强大的框架，专为研究人员和开发者设计，旨在助力大家快速构建属于自己的科学智能体（Scientific Agents），免受工具调用、技能组合、记忆存储等工程化工作的烦扰。\n\n**[MagiClaw](https://github.com/sjtu-sai-agents/MagiClaw)** 是基于 EvoMaster 开发的飞书智能体助手。通过自然语言对话，它能帮你基于 EvoMaster 框架创建新智能体，或调度多个已有的智能体协作完成任务。\n\n## \u003ca id=\"key-features\"\u003e\u003c/a\u003e✨ 核心特性\n\n### 1. ♾️ 通用兼容性\n\nEvoMaster 支持并适配当前智能体领域的主流技术栈，无论是 [MCP](https://www.anthropic.com/news/model-context-protocol) 工具调用还是[技能（Skills）](https://platform.claude.com/docs/en/agents-and-tools/agent-skills/overview)，都可以通过一行配置代码接入你的智能体。\n\n### 2. ⚡ 极速开发\n\nEvoMaster 的设计理念是便携与易用，通过即插即用的组件设计让你无需重写核心逻辑即可快速上手进行开发。仅需 **约 100 行代码** 即可启动一个自定义智能体——代码复杂度不应成为创新的阻碍。\n\n### 3. 🧬 智能体自主进化\n\n通过 EvoMaster 实现的 [MagiClaw](https://github.com/sjtu-sai-agents/MagiClaw)，不仅能让使用者通过自然语言调度多个已有的智能体协作完成任务，还能继续基于 EvoMaster 框架创建新智能体，实现智能体的自我迭代与进化。\n\n### 4. 🔁 运行级自进化\n\nEvoMaster 可以可选地先运行智能体，再分析其轨迹和环境反馈，生成 run-local skills, tools和prompt overlays，随后自动用进化后的 overlay 再运行同一个智能体。工具改进会以可 review 的 proposal 形式保存。详情请见[自进化使用指南](./docs/zh/evolution.md)。\n\n### \u003ca id=\"scimaster-series\"\u003e\u003c/a\u003e5. 🔬 SciMaster 生态系统\n\n我们基于 EvoMaster 统一实现并开源了多个 SciMaster 系列智能体。您可以快速部署 SciMaster 系列的成熟智能体，也可以轻松将其迁移至生物学、材料科学等其他科学领域。\n\n| 智能体名称 | 领域 / 专长 | 论文 / 链接 | 状态 |\n| --- | --- | --- | --- |\n| **ML-Master 2.0** | 自主机器学习 (Autonomous Machine Learning) | [ArXiv:2601.10402](https://arxiv.org/abs/2601.10402) | 可用 |\n| **ML-Master** | 自主机器学习 (Autonomous Machine Learning) | [ArXiv:2506.16499](https://arxiv.org/abs/2506.16499) | 可用 |\n| **X-Master** | 通用科学智能体 (General Scientific Agent) | [ArXiv:2507.05241](https://arxiv.org/abs/2507.05241) | 可用 |\n| **Browse-Master** | 网页搜索智能体 (Web Search Agent) | [ArXiv:2508.09129](https://arxiv.org/abs/2508.09129) | 可用 |\n| **PhysMaster** | 物理研究与推理 (Physics Research \u0026 Reasoning) | [ArXiv:2512.19799](https://arxiv.org/abs/2512.19799) | 敬请期待 |\n| **EmboMaster** | 具身智能训练 (Embodied Intelligence Training) | [ArXiv:2601.21570](https://arxiv.org/abs/2601.21570) | 敬请期待 |\n\n（更多 SciMaster 系列智能体敬请期待...）\n\n---\n\n## \u003ca id=\"roadmap\"\u003e\u003c/a\u003e🗺️ 路线图\n\n| 阶段 | 版本 | 内容 | 状态 |\n|------|------|------|------|\n| **当前** | v0.0.x | 核心框架、基础文档、简易智能体示例 | ✅ 已完成 |\n| **第一阶段** | v0.1.x | SciMaster 系列智能体实现开源 | ✅ 已完成 |\n| **第二阶段** | v0.2.x | [MagiClaw](https://github.com/sjtu-sai-agents/MagiClaw) 飞书智能体助手开源 | ✅ 已完成 |\n| **第三阶段** | v0.3.x | Bohrium 工具库 — 集成 [Bohrium](https://www.bohrium.com/)，原生支持 30,000+ 科学工具与 API | 💡 探索中 |\n\n\n\n## 🏗️ 项目架构\n\n```\nEvoMaster/\n├── evomaster/                        # 核心库\n│   ├── agent/                        # Agent 组件（Agent, Session, Tools）\n│   ├── core/                         # 工作流（Exp, Playground）\n│   ├── env/                          # 环境（Docker, Local）\n│   ├── skills/                       # 技能系统\n│   ├── skills_ts/                    # TypeScript 技能（OpenClaw bridge）\n│   └── utils/                        # 工具（LLM, Types）\n├── extensions/                       # 通过skill使用EvoMaster及官方示例自定义工具\n├── playground/                       # Playground 实现\n├── configs/                          # 配置文件\n└── docs/                             # 文档\n```\n\n## 📚 文档\n\n完整文档目录请参阅 [docs/README_zh.md](./docs/README_zh.md)。\n\n\n## \u003ca id=\"quick-start\"\u003e\u003c/a\u003e🚀 快速开始\n\n### 📦 安装\n\n#### 使用 pip\n\n```bash\n# 克隆仓库\ngit clone -b main --single-branch https://github.com/sjtu-sai-agents/EvoMaster.git\ncd EvoMaster\n\n# 安装依赖\npip install -r requirements.txt\n\n# 在 configs/ 中配置 LLM API 密钥\n```\n\n#### 使用 uv\n\n[uv](https://docs.astral.sh/uv/) 是一个快速的 Python 包安装器。可以使用以下任一方式：\n\n```bash\n# 选项 1：从 pyproject.toml + uv.lock 同步（推荐）\nuv sync\n\n# 选项 2：从 requirements.txt 安装\nuv pip install -r requirements.txt\n```\n\n创建虚拟环境并使用 uv 运行：`uv venv \u0026\u0026 source .venv/Scripts/activate`（Windows）或 `source .venv/bin/activate`（Linux/macOS），然后运行 `uv sync`。\n\n### 使用您的 API Key\n\n打开位于 `configs/[playground name]` 的配置文件并填写相应的空白处。例如，如果您想使用 Deepseek-V3.2 运行 `minimal_multi_agent`，请打开 `configs/minimal_multi_agent/deepseek-v3.2-example.yaml` 并修改如下内容：\n\n```bash\n  local_sglang:\n    provider: \"deepseek\"\n    model: \"deepseek-v3.2\"\n    api_key: \"dummy\"\n    base_url: \"http://192.168.2.110:18889/v1\"\n```\n\n如果您的模型 API 支持 OpenAI 格式，也可以使用 `openai` 配置。请记得同时修改后续 Agent 的 LLM 配置。\n\n### 使用环境变量 (.env)\n\n您也可以使用环境变量进行配置。这种方式更加安全和灵活：\n\n1. **从模板创建 `.env` 文件：**\n   ```bash\n   cp .env.template .env\n   ```\n\n2. **编辑 `.env` 文件**并填写您的 API 密钥和配置值：\n   ```bash\n   # 示例：设置您的 DeepSeek API 密钥\n   DEEPSEEK_API_KEY=\"your-api-key-here\"\n   DEEPSEEK_API_BASE=\"http://127.0.0.1:18889/v1\"\n   ```\n\n3. **运行您的命令：**\n\n   系统会自动从项目根目录加载 `.env` 文件，因此您可以直接运行：\n   ```bash\n   python run.py --agent minimal --task \"你的任务描述\"\n   ```\n\n   或者，您也可以使用 `dotenv` CLI 工具：\n   ```bash\n   dotenv run python run.py --agent minimal --task \"你的任务描述\"\n   ```\n\n### 基本使用\n\n```bash\ncd EvoMaster\npython run.py --agent minimal --task \"你的任务描述\"\n```\n\n### 使用自定义配置\n\n```bash\npython run.py --agent minimal --config configs/minimal/config.yaml --task \"你的任务描述\"\n```\n\n### 从文件读取任务\n\n```bash\npython run.py --agent minimal --task task.txt\n```\n\n## 📋 示例\n关于 Playground 示例的详情请参阅[这里](./playground/README_CN.md)。以下示例都使用本地环境运行代码。同时，EvoMaster 也支持在 Docker 中安全运行代码，仅需简单几步即可完成，请参阅[这里](./docs/zh/docker_session.md)了解更多。\n\n### 单智能体（Minimal）\n```bash\npython run.py --agent minimal --config configs/minimal/deepseek-v3.2-example.yaml --task \"Discover a pattern: Given sequence 1, 4, 9, 16, 25... find the formula\"\n```\n\n### 自进化智能体（Minimal）\n先运行 baseline，再从轨迹中生成 run-local skills 和 prompt overlays，最后用进化后的 overlay 重新运行。详见[自进化使用指南](./docs/zh/evolution.md)。\n\n```bash\npython run.py \\\n  --agent minimal \\\n  --config configs/minimal/gpt-5-example.yaml \\\n  --task \"Discover a pattern: Given sequence 1, 4, 9, 16, 25... find the formula\" \\\n  --run-dir runs/evo_test \\\n  --evolve \\\n  --evolve-iterations 2\n```\n\n### 单智能体，输入任务包含图片（Minimal）\n```bash\npython run.py --agent minimal --config configs/minimal/deepseek-v3.2-example.yaml --task \"Describe what you see in these images\" --images /path/to/image1.png /path/to/image2.jpg\n```\n\n### 单智能体使用 TypeScript 格式的 Skill\n```bash\npython run.py --agent minimal_openclaw_skill --config configs/minimal_openclaw_skill/config.yaml --task \"总结这个飞书文档的内容 \u003c你的飞书文档网址\u003e\"\n```\n\n### 玻尔（Bohrium）平台科学计算工具\n请参考 [minimal_bohrium README](./playground/minimal_bohrium/README_CN.md)\n\n### 简单多智能体系统\n```bash\npython run.py --agent minimal_multi_agent --config configs/minimal_multi_agent/deepseek-v3.2-example.yaml --task \"Write a Python program that implements the following features: Read a text file (create a sample file if it doesn't exist). Count the occurrences of each word in the file. Sort the results by frequency in descending order. Save the results to a new file named word_count.txt. Output the top 10 most common words to the terminal.\"\n```\n\n\n### 多智能体系统（Exp 级并行）\n```bash\npython run.py --agent minimal_multi_agent_parallel --config configs/minimal_multi_agent_parallel/deepseek-v3.2-example.yaml --task \"Write a Python program that implements the following features: Read a text file (create a sample file if it doesn't exist). Count the occurrences of each word in the file. Sort the results by frequency in descending order. Save the results to a new file named word_count.txt. Output the top 10 most common words to the terminal.\"\n```\n\n### Kaggle 自动化\n```bash\npip install -r playground/minimal_kaggle/requirements.txt\npython run.py --agent minimal_kaggle --config configs/minimal_kaggle/deepseek-v3.2-example.yaml --task playground/minimal_kaggle/data/public/description.md\n```\n\n\n### X-Master 工作流\n更多详情请参阅 [X-Master README](./playground/x_master/README_CN.md)\n```bash\n# 安装 mcp_sandbox 环境\npip install -r playground/x_master/mcp_sandbox/requirements.txt\npython run.py --agent x_master --task \"Which condition of Arrhenius's sixth impossibility theorem do critical-level views violate?\\n\\nAnswer Choices:\\nA. Egalitarian Dominance\\nB. General Non-Extreme Priority\\nC. Non-Elitism\\nD. Weak Non-Sadism\\nE. Weak Quality Addition\"\n```\n\n### ML-Master 1.0\n更多详情请参阅 [ML-Master 1.0 README](./playground/ml_master/README_CN.md)\n```bash\npip install -r playground/ml_master/requirements.txt\npython run.py --agent ml_master --config configs/ml_master/config.yaml --task /data/exp_data/detecting-insults-in-social-commentary/prepared/public/description.md\n```\n\n### ML-Master 2.0\n更多详情请参阅 [ML-Master 2.0 README](./playground/ml_master_2/README_CN.md)\n```bash\npip install -r playground/ml_master_2/requirements.txt\n# 可选\n# export HF_ENDPOINT=https://hf-mirror.com\npython run.py --agent ml_master_2 --config configs/ml_master_2/deepseek-v3.2-example.yaml --task playground/ml_master_2/data/detecting-insults-in-social-commentary/prepared/public/description.md\n```\n\n### Browse-Master 工作流\n更多详情请参阅 [Browse-Master README](./playground/browse_master/README_CN.md)\n```bash\n# 安装 mcp_sandbox 环境\npip install -r playground/browse_master/mcp_sandbox/requirements.txt\npython run.py --agent browse_master --config configs/browse_master/config.yaml --task \"I am searching for the pseudonym of a writer and biographer who authored numerous books, including their autobiography. In 1980, they also wrote a biography of their father. The writer fell in love with the brother of a philosopher who was the eighth child in their family. The writer was divorced and remarried in the 1940s.\"\n```\n\n## 🤝 参与贡献\n欢迎为 EvoMaster 做出贡献！如果您有任何想法、bug 修复或新特性，欢迎提交 Pull Request。如果涉及较大的变更，请先提交 issue 与我们讨论您的改动方案。\n\n## \u003ca id=\"contact-us\"\u003e\u003c/a\u003e💬 联系我们\n\n我们欢迎讨论，问题和反馈! 欢迎加入我们微信社群:\n\n\u003cdiv align=\"center\"\u003e\n\n\u003cimg src=\"https://github.com/user-attachments/assets/edb4d2bd-4012-40ce-b73c-2a9d75be41c9\" width=\"250\" alt=\"WeChat Group QR Code\"/\u003e\n\n*扫描二维码加入微信群*\n\n\u003c/div\u003e\n\n## ✍️ 引用\n\n如果认为这个工作对您有帮助，请引用以下文章。\n\n```bibtex\n@misc{zhu2026evomasterfoundationalagentframework,\n      title={EvoMaster: A Foundational Agent Framework for Building Evolving Autonomous Scientific Agents at Scale}, \n      author={Xinyu Zhu and Yuzhu Cai and Zexi Liu and Cheng Wang and Fengyang Li and Wenkai Jin and Wanxu Liu and Zehao Bing and Bingyang Zheng and Jingyi Chai and Shuo Tang and Rui Ye and Yuwen Du and Xianghe Pang and Yaxin Du and Tingjia Miao and Yuzhi Zhang and Ruoxue Liao and Zhaohan Ding and Linfeng Zhang and Yanfeng Wang and Weinan E and Siheng Chen},\n      year={2026},\n      eprint={2604.17406},\n      archivePrefix={arXiv},\n      primaryClass={cs.AI},\n      url={https://arxiv.org/abs/2604.17406}, \n}\n```\n\n## ⭐ Star 记录\n如果您觉得 EvoMaster 和 MagiClaw 对您有帮助，请为我们点一个 Star 支持！⭐\n\u003ca href=\"https://www.star-history.com/?repos=sjtu-sai-agents%2FEvoMaster\u0026type=date\u0026legend=top-left\"\u003e\n \u003cpicture\u003e\n   \u003csource media=\"(prefers-color-scheme: dark)\" srcset=\"https://api.star-history.com/image?repos=sjtu-sai-agents/EvoMaster\u0026type=date\u0026theme=dark\u0026legend=top-left\" /\u003e\n   \u003csource media=\"(prefers-color-scheme: light)\" srcset=\"https://api.star-history.com/image?repos=sjtu-sai-agents/EvoMaster\u0026type=date\u0026legend=top-left\" /\u003e\n   \u003cimg alt=\"Star History Chart\" src=\"https://api.star-history.com/image?repos=sjtu-sai-agents/EvoMaster\u0026type=date\u0026legend=top-left\" /\u003e\n \u003c/picture\u003e\n\u003c/a\u003e\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fsjtu-sai-agents%2FEvoMaster","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fsjtu-sai-agents%2FEvoMaster","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fsjtu-sai-agents%2FEvoMaster/lists"}