[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"tool-berylliumsec--nebula":3,"similar-berylliumsec--nebula":203},{"id":4,"github_repo":5,"name":6,"description_en":7,"description_zh":8,"ai_summary_zh":8,"readme_en":9,"readme_zh":10,"quickstart_zh":11,"use_case_zh":12,"hero_image_url":13,"owner_login":14,"owner_name":14,"owner_avatar_url":15,"owner_bio":16,"owner_company":17,"owner_location":17,"owner_email":17,"owner_twitter":17,"owner_website":17,"owner_url":18,"languages":19,"stars":40,"forks":41,"last_commit_at":42,"license":43,"difficulty_score":44,"env_os":45,"env_gpu":46,"env_ram":47,"env_deps":48,"category_tags":54,"github_topics":59,"view_count":75,"oss_zip_url":17,"oss_zip_packed_at":17,"status":76,"created_at":77,"updated_at":78,"faqs":79,"releases":115},9251,"berylliumsec\u002Fnebula","nebula","AI-powered penetration testing assistant for automating recon, note-taking, and vulnerability analysis.","Nebula 是一款专为网络安全专业人士和道德黑客打造的 AI 渗透测试助手。它将先进的大语言模型（如 Llama 3.1、Mistral 及 OpenAI 系列）直接集成到命令行界面中，旨在解决传统渗透测试中信息收集繁琐、笔记整理耗时以及漏洞分析依赖人工经验等痛点。\n\n通过 Nebula，用户可以实现工作流程的自动化与智能化：它能根据终端工具的输出实时提供漏洞发现与利用建议，自动记录并分类安全发现，甚至支持联网搜索最新的网络安全动态以辅助决策。此外，Nebula 还内置了截图标注和手动笔记功能，确保测试过程文档化的完整与高效。\n\n该工具特别适合需要提升效率的安全研究员、渗透测试工程师以及希望将 AI 能力融入现有 CLI 工作流的开发者。其独特亮点在于灵活的模型支持，既可通过 Ollama 本地部署开源模型以保障数据隐私，也能调用云端 API 获取更强算力；同时，它不仅能理解自然语言指令，还能作为智能代理协调外部工具，让复杂的攻防演练变得更加直观和流畅。","# Nebula – AI-Powered Penetration Testing Assistant\n\nNebula is an advanced, AI-powered penetration testing open-source tool that revolutionizes penetration testing by integrating state-of-the-art AI models into your command-line interface. Designed for cybersecurity professionals, ethical hackers, and developers, Nebula automates vulnerability assessments and enhances security workflows with real-time insights and automated note-taking.\n\n\n![Nebula AI-Powered Penetration Testing CLI Interface](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fberylliumsec_nebula_readme_bf3daa592b95.png)\n\n## Acknowledgement\n\n**First i would like to thank the All-Mighty God who is the source of all knowledge, without Him, this would not be possible.**\n\n## News\n\nIntroducing the Deep Application Profiler (DAP). DAP uses neural networks to analyze an executable's internal structure and intent, rather than relying on traditional virus signatures. This approach enables it to detect new, zero-day malware that conventional methods often miss. DAP also provides detailed breakdowns for rapid analyst review and is available as both a web service and an API. [Learn More Here](https:\u002F\u002Fberylliumsec.com\u002Fmalware-analysis)\n\n\n## Nebula: AI-Powered Penetration Testing Platform\n\nNebula is a cutting-edge, AI-powered penetration testing tool designed for cybersecurity professionals and ethical hackers. It integrates advanced open-source AI models such as OpenAI's models (any model that is available via API) Meta's Llama-3.1-8B-Instruct, Mistralai's Mistral-7B-Instruct-v0.2, and DeepSeek-R1-Distill-Llama-8B—directly into the command line interface (CLI). By leveraging these state-of-the-art models, Nebula not only enhances vulnerability assessments and penetration testing workflows but also supports any tool that can be invoked from the CLI.\n\n\n## Installation\n\n**System Requirements:**\n\nFor CPU-Based Inference(Ollama)(Note that Ollama Supports GPU too):\n- At least 16GB of RAM \n- Python 3.10 – 3.13.9\n- [Ollama](https:\u002F\u002Follama.com\u002F)\n\n**Installation Command:**\n```bash\npython -m pip install nebula-ai --upgrade\n```\n\n\n## Running Nebula\n\n**Important:** \n\n\n**Ollama Local Model Based Usage**\n\n[Install Ollama](https:\u002F\u002Follama.com\u002Fdownload\u002Fmac) and download your preferred models for example\n\n```bash\n ollama pull mistral\n```\nThen enter the model's exact name as it appears in Ollama in the engagement settings.\n\n**OpenAI Models Usage**\n\nTo use OpenAI models, add your API keys to your env like so\n\n```bash\nexport OPENAI_API_KEY=\"sk-blah-blaj\"\n```\n\nThen enter the OpenAI model's exact name in the engagement settings.\n\n\nRun nebula\n\n```\nnebula\n```\n\n**Using docker**\n\nFirst allow local connections to your X server:\n\n```bash\nxhost +local:docker\n```\n\n```bash\ndocker run --rm -it   -e DISPLAY=$DISPLAY   -v \u002Fhome\u002FYOUR_HOST_NAME\u002F.local\u002Fshare\u002Fnebula\u002Flogs:\u002Froot\u002F.local\u002Fshare\u002Fnebula\u002Flogs -v YOUR_ENGAGEMENT_FOLDER_ON_HOST_MACHINE:\u002Fengagements -v \u002Ftmp\u002F.X11-unix:\u002Ftmp\u002F.X11-unix   berylliumsec\u002Fnebula:latest\n```\n### Interacting with the models. \n\nTo interact with the models, begin your input with a `!` or use the AI\u002FTerminal button to switch between modes. For example: `! write a python script to scan the ports of a remote system` the \"!\" is not needed if you use the context button\n\n## Key Features\n\n- **AI-Powered Internet Search via agents:**  \n  Enhance responses by integrating real-time, internet-sourced context to keep you updated on cybersecurity trends. \"whats in the news on cybersecurity today\"\n  \n- **AI-Assisted Note-Taking:**  \n  Automatically record and categorize security findings.\n\n- **Real-Time AI-Driven Insights:**  \n  Get immediate suggestions for discovering and exploiting vulnerabilities based on terminal tool outputs.\n\n- **Enhanced Tool Integration:**  \n  Seamlessly import data from external tools for AI-powered note-taking and advice.\n\n- **Integrated Screenshot & Editing:**  \n  Capture and annotate images directly within Nebula for streamlined documentation.\n\n- **Manual Note-Taking & Automatic Command Logging:**  \n  Maintain a detailed log of your actions and findings with both automated and manual note-taking features.\n  \n- **Status feed:**  \n  This panel displays your most recent penetration testing activities, it refreshes every five minutes\n\n\n### Roadmap\n\n- Create custom models that are more useful for penetration testing\n\n### Troubleshooting\n\nLogs are located at `\u002Fhome\u002F[your_username]\u002F.local\u002Fshare\u002Fnebula\u002Flogs`. You would most likely find the reason for the error in one of those logs\n\n## Get More Support\n\n- Have questions or need help? [Open an Issue](https:\u002F\u002Fgithub.com\u002Fberylliumsec\u002Fnebula\u002Fissues) on GitHub.\n","# Nebula – 基于AI的渗透测试助手\n\nNebula 是一款先进的、基于 AI 的开源渗透测试工具，通过将最前沿的 AI 模型集成到您的命令行界面中，彻底革新了渗透测试流程。Nebula 专为网络安全专业人士、道德黑客和开发者设计，能够自动化漏洞评估，并借助实时洞察和自动化的笔记记录功能，提升安全工作流效率。\n\n\n![Nebula 基于AI的渗透测试 CLI 界面](https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fberylliumsec_nebula_readme_bf3daa592b95.png)\n\n## 致谢\n\n**首先，我要感谢全能的上帝，祂是一切知识的源泉，若没有祂，这一切都不可能实现。**\n\n## 新闻\n\n隆重推出深度应用分析器（DAP）。DAP 利用神经网络来分析可执行文件的内部结构和意图，而不依赖传统的病毒特征库。这种方法使其能够检测到传统方法常常遗漏的新型零日恶意软件。DAP 还提供详细的分解信息，便于分析师快速审查，并且既可作为 Web 服务使用，也可通过 API 调用。[了解详情请点击这里](https:\u002F\u002Fberylliumsec.com\u002Fmalware-analysis)\n\n\n## Nebula：基于 AI 的渗透测试平台\n\nNebula 是一款尖端的、基于 AI 的渗透测试工具，专为网络安全专业人士和道德黑客打造。它将 OpenAI 的各类可通过 API 调用的模型、Meta 的 Llama-3.1-8B-Instruct、Mistral 的 Mistral-7B-Instruct-v0.2 以及 DeepSeek-R1-Distill-Llama-8B 等先进的开源 AI 模型直接集成到命令行界面（CLI）中。借助这些最先进的模型，Nebula 不仅能够增强漏洞评估和渗透测试的工作流，还支持任何可通过 CLI 调用的工具。\n\n\n## 安装\n\n**系统要求：**\n\n对于基于 CPU 的推理（Ollama）（注意：Ollama 同样支持 GPU）：\n- 至少 16GB 内存\n- Python 3.10 – 3.13.9\n- [Ollama](https:\u002F\u002Follama.com\u002F)\n\n**安装命令：**\n```bash\npython -m pip install nebula-ai --upgrade\n```\n\n\n## 运行 Nebula\n\n**重要提示：** \n\n\n**基于 Ollama 本地模型的使用**\n\n请先安装 Ollama（[下载地址](https:\u002F\u002Follama.com\u002Fdownload\u002Fmac)），并下载您偏好的模型，例如：\n\n```bash\n ollama pull mistral\n```\n然后在项目设置中输入该模型在 Ollama 中显示的确切名称。\n\n**OpenAI 模型的使用**\n\n要使用 OpenAI 模型，请将您的 API 密钥添加到环境变量中，如下所示：\n\n```bash\nexport OPENAI_API_KEY=\"sk-blah-blaj\"\n```\n\n随后，在项目设置中输入 OpenAI 模型的确切名称。\n\n\n运行 Nebula：\n\n```bash\nnebula\n```\n\n\n**使用 Docker**\n\n首先允许本地连接到您的 X 服务器：\n\n```bash\nxhost +local:docker\n```\n\n然后运行以下命令：\n\n```bash\ndocker run --rm -it   -e DISPLAY=$DISPLAY   -v \u002Fhome\u002FYOUR_HOST_NAME\u002F.local\u002Fshare\u002Fnebula\u002Flogs:\u002Froot\u002F.local\u002Fshare\u002Fnebula\u002Flogs -v YOUR_ENGAGEMENT_FOLDER_ON_HOST_MACHINE:\u002Fengagements -v \u002Ftmp\u002F.X11-unix:\u002Ftmp\u002F.X11-unix   berylliumsec\u002Fnebula:latest\n```\n\n### 与模型交互\n\n要与模型交互，只需在输入前加上 `!`，或使用 AI\u002F终端按钮切换模式。例如：`! 编写一个 Python 脚本来扫描远程系统的端口`。如果您使用上下文按钮，则无需添加 `!`。\n\n## 核心功能\n\n- **基于代理的 AI 驱动互联网搜索：**  \n  通过整合实时的互联网来源上下文，帮助您及时了解最新的网络安全动态，从而丰富响应内容。“今天网络安全领域有哪些新闻？”\n\n- **AI 辅助笔记记录：**  \n  自动记录并分类安全发现结果。\n\n- **实时 AI 驱动洞察：**  \n  根据终端工具的输出，即时为您提供发现和利用漏洞的建议。\n\n- **增强的工具集成：**  \n  无缝导入外部工具的数据，用于 AI 驱动的笔记记录和建议。\n\n- **集成截图与编辑功能：**  \n  在 Nebula 内直接捕获并标注图像，简化文档编写流程。\n\n- **手动笔记记录与自动命令日志：**  \n  结合自动和手动笔记功能，详细记录您的操作和发现。\n\n- **状态信息流：**  \n  此面板会显示您最近的渗透测试活动，每五分钟刷新一次。\n\n\n### 路线图\n\n- 构建更适用于渗透测试的自定义模型\n\n### 故障排除\n\n日志文件位于 `\u002Fhome\u002F[your_username]\u002F.local\u002Fshare\u002Fnebula\u002Flogs`。您很可能在这些日志中找到问题的原因。\n\n## 获取更多支持\n\n- 如有任何疑问或需要帮助，请在 GitHub 上 [提交一个问题](https:\u002F\u002Fgithub.com\u002Fberylliumsec\u002Fnebula\u002Fissues)。","# Nebula AI 渗透测试助手快速上手指南\n\nNebula 是一款先进的开源 AI 渗透测试工具，它将前沿的大语言模型（如 Llama 3.1、Mistral、DeepSeek 及 OpenAI 系列）集成到命令行界面（CLI）中。它能自动化漏洞评估、提供实时安全洞察并自动记录笔记，专为网络安全专业人员、白帽黑客和开发者设计。\n\n## 环境准备\n\n在开始之前，请确保您的系统满足以下要求：\n\n*   **操作系统**：支持 Linux 或 macOS（Windows 用户建议使用 WSL2 或 Docker）。\n*   **Python 版本**：Python 3.10 – 3.13.9。\n*   **内存要求**：若使用本地模型推理，建议至少 **16GB RAM**。\n*   **前置依赖**：\n    *   **本地模型方案**：需安装 [Ollama](https:\u002F\u002Follama.com\u002F)（支持 CPU\u002FGPU 加速）。\n    *   **云端模型方案**：需拥有 OpenAI API Key 或其他兼容 API 的密钥。\n\n> **提示**：国内用户在安装 Ollama 或拉取大模型时，若遇到网络延迟，可配置相关代理或使用国内镜像源加速。\n\n## 安装步骤\n\n### 1. 安装核心工具\n使用 pip 安装 Nebula：\n\n```bash\npython -m pip install nebula-ai --upgrade\n```\n\n### 2. 配置 AI 模型\n\n根据您的使用场景选择以下一种方式进行配置：\n\n#### 方案 A：使用本地模型 (推荐，无需联网即可推理)\n1.  安装并启动 [Ollama](https:\u002F\u002Follama.com\u002Fdownload)。\n2.  拉取您偏好的模型（例如 Mistral）：\n    ```bash\n    ollama pull mistral\n    ```\n    *(注：也可替换为 `llama3.1`、`deepseek-r1` 等 Ollama 支持的模型)*\n3.  启动 Nebula 后，在交互设置中输入与 Ollama 中完全一致的模型名称。\n\n#### 方案 B：使用 OpenAI 系列模型\n1.  将您的 API Key 导出到环境变量：\n    ```bash\n    export OPENAI_API_KEY=\"sk-your-actual-api-key\"\n    ```\n2.  启动 Nebula 后，在交互设置中输入对应的 OpenAI 模型名称（如 `gpt-4o`）。\n\n## 基本使用\n\n### 启动工具\n在终端直接输入以下命令启动 Nebula：\n\n```bash\nnebula\n```\n\n### 与 AI 交互\nNebula 提供两种主要的交互模式：\n\n1.  **命令前缀模式**：在输入内容前加上 `!` 即可触发 AI 处理。\n    *   示例：让 AI 编写一个端口扫描脚本\n        ```text\n        ! write a python script to scan the ports of a remote system\n        ```\n\n2.  **模式切换**：使用界面中的 **AI\u002FTerminal** 按钮或 **Context** 按钮在纯终端模式和 AI 辅助模式间切换。使用 Context 按钮时，无需输入 `!` 前缀。\n\n### 核心功能速览\n*   **实时情报搜索**：询问 `\"whats in the news on cybersecurity today\"` 可获取最新的网络安全趋势。\n*   **自动笔记**：工具会自动分类记录您的安全发现和操作日志。\n*   **截图与标注**：直接在界面内捕获屏幕并进行标注，便于报告撰写。\n*   **状态反馈**：侧边栏每 5 分钟刷新一次，显示最近的渗透测试活动状态。\n\n> **注意**：所有操作日志默认保存在 `~\u002F.local\u002Fshare\u002Fnebula\u002Flogs` 目录下，如遇错误请优先检查此处日志。","某安全团队正在对一家金融客户的内部网络进行渗透测试，需要在有限时间内完成从信息收集到漏洞分析的全流程。\n\n### 没有 nebula 时\n- 测试人员需手动在多个终端窗口切换，运行 Nmap、Nikto 等工具后，再人工复制输出结果到笔记软件中整理，效率极低且容易遗漏关键信息。\n- 面对海量的扫描日志，分析师必须依靠个人经验逐行排查潜在漏洞，缺乏实时的智能建议，导致零日漏洞或隐蔽弱点容易被忽视。\n- 编写用于特定环境验证的 Python 脚本或 Exploit 代码需要中断测试流程去查阅文档或搜索示例，严重拖慢进攻节奏。\n- 最终报告撰写阶段，需要花费大量时间回溯操作历史来拼凑测试路径，截图和标注工作繁琐，难以保证文档的专业性和完整性。\n\n### 使用 nebula 后\n- 直接在 CLI 界面中调用 nebula，它自动执行侦察命令并实时分类记录发现，将分散的工具输出整合为结构化的安全笔记，无需手动复制粘贴。\n- 基于集成的 Llama-3.1 或 Mistral 等模型，nebula 能即时分析终端输出，主动提示潜在的利用路径和未察觉的风险点，显著提升漏洞检出率。\n- 只需输入\"!write a python script to scan ports\"，nebula 即可现场生成并解释定制化脚本，让测试人员无需离开当前上下文即可快速验证假设。\n- 内置的截图与编辑功能允许随时捕获证据并自动关联到对应笔记，结合自动命令日志，一键生成详尽且逻辑清晰的审计报告草稿。\n\nnebula 通过将 AI 深度融入命令行工作流，把渗透测试从繁琐的手工劳动转变为智能化的决策辅助过程，极大提升了攻防演练的效率与深度。","https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fberylliumsec_nebula_0296a39f.png","berylliumsec","https:\u002F\u002Foss.gittoolsai.com\u002Favatars\u002Fberylliumsec_fa58194e.png","",null,"https:\u002F\u002Fgithub.com\u002Fberylliumsec",[20,24,28,32,36],{"name":21,"color":22,"percentage":23},"Python","#3572A5",96,{"name":25,"color":26,"percentage":27},"HTML","#e34c26",1.8,{"name":29,"color":30,"percentage":31},"CSS","#663399",1.7,{"name":33,"color":34,"percentage":35},"Dockerfile","#384d54",0.5,{"name":37,"color":38,"percentage":39},"Shell","#89e051",0.1,932,139,"2026-04-15T11:21:31","BSD-2-Clause",3,"Linux, macOS","非必需。支持 CPU 推理；若使用 Ollama 则可选支持 GPU（具体型号和显存未说明）","最低 16GB",{"notes":49,"python":50,"dependencies":51},"主要依赖 Ollama 运行本地模型（如 Mistral、Llama-3.1 等），也支持通过 API 调用 OpenAI 模型。若使用 Docker 运行，需配置 X Server 以支持图形界面（执行 xhost +local:docker）。日志默认存储在 ~\u002F.local\u002Fshare\u002Fnebula\u002Flogs。","3.10 - 3.13.9",[52,53],"ollama","nebula-ai",[55,56,57,58],"开发框架","图像","Agent","语言模型",[60,61,62,63,64,65,66,67,68,69,70,71,72,73,74],"ethical-hacking-tool","penetration-testing-framework","penetration-testing-tool","ai-powered-ethical-hacking-tool","ai-powered-penetration-testing-tool","ai","python","cybersecurity","cybersecurity-tools","ethical-hacking","llm","security-automation","vulnerability-assesment-tools","vulnerability-assessment","vulnerability-scanners",2,"ready","2026-03-27T02:49:30.150509","2026-04-19T03:05:57.285774",[80,85,90,95,100,105,110],{"id":81,"question_zh":82,"answer_zh":83,"source_url":84},41523,"为什么在 Kali Linux 上运行 pip install --upgrade 无法升级到最新版本？","这是因为旧版本将允许的最大 Python 版本限制为 3.11.x，而新版 Kali Linux 默认运行 Python 3.12.x。该问题已在最新发行版中修复，请再次运行升级命令即可正常更新。","https:\u002F\u002Fgithub.com\u002Fberylliumsec\u002Fnebula\u002Fissues\u002F97",{"id":86,"question_zh":87,"answer_zh":88,"source_url":89},41524,"运行 Nebula 加载 Nuclei 模型时程序被杀死或设备冻结怎么办？","这通常是因为系统内存不足。维护者已添加支持一次只加载一个模型的功能以减少内存占用。请升级到最新版本再试。如果问题依旧，请检查您的系统是否满足 README 中的内存要求。","https:\u002F\u002Fgithub.com\u002Fberylliumsec\u002Fnebula\u002Fissues\u002F33",{"id":91,"question_zh":92,"answer_zh":93,"source_url":94},41525,"使用 pip 安装运行时出现 FileNotFoundError: No such file or directory: '.\u002Funified_models'错误如何解决？","该问题已在代码提交中修复。请运行以下命令获取包含修复的最新版本：\n1. pip install nebula-ai --upgrade\n如果是 Docker 用户，请拉取最新镜像：docker pull berylliumsec\u002Fnebula:latest","https:\u002F\u002Fgithub.com\u002Fberylliumsec\u002Fnebula\u002Fissues\u002F8",{"id":96,"question_zh":97,"answer_zh":98,"source_url":99},41526,"如何在 Android 设备上的容器化 Kali (NetHunter) 中运行 Nebula？","可以在 Android 设备的容器化 Kali 环境中运行，但需注意这可能是一个新手容易遇到配置问题的场景。建议同时尝试在桌面版 Kali 上运行以获得更强大的性能和图形界面支持。如果在移动端遇到问题，请确保未破坏 Kali 安装，并尝试使用 pipx 简化安装过程。","https:\u002F\u002Fgithub.com\u002Fberylliumsec\u002Fnebula\u002Fissues\u002F91",{"id":101,"question_zh":102,"answer_zh":103,"source_url":104},41527,"如何申请试用或演示 Nebula Pro 版本？","请访问官方等待列表表单进行申请：https:\u002F\u002Fwww.berylliumsec.com\u002Fnebula-pro-waitlist","https:\u002F\u002Fgithub.com\u002Fberylliumsec\u002Fnebula\u002Fissues\u002F103",{"id":106,"question_zh":107,"answer_zh":108,"source_url":109},41528,"界面提示建议功能不存在或无法输入问题怎么办？","请升级到最新版本以解决该问题。","https:\u002F\u002Fgithub.com\u002Fberylliumsec\u002Fnebula\u002Fissues\u002F66",{"id":111,"question_zh":112,"answer_zh":113,"source_url":114},41529,"搜索结果的排序顺序是怎样的？为什么最近运行的命令没有显示在最后？","之前的版本中结果排序可能存在异常。最新发行版已经改进了搜索结果的显示和排序逻辑，请升级到最新版本以获得正确的排序体验。","https:\u002F\u002Fgithub.com\u002Fberylliumsec\u002Fnebula\u002Fissues\u002F39",[116,121,125,129,133,137,141,145,149,153,157,162,166,170,175,180,185,190,195,199],{"id":117,"version":118,"summary_zh":119,"released_at":120},333489,"2.0.0b31","发布说明：\n   - 支持较旧版本的 Python","2026-04-06T16:07:49",{"id":122,"version":123,"summary_zh":119,"released_at":124},333490,"2.0.0b30","2026-04-06T15:41:15",{"id":126,"version":127,"summary_zh":119,"released_at":128},333491,"2.0.0b29","2026-04-06T15:40:05",{"id":130,"version":131,"summary_zh":119,"released_at":132},333492,"2.0.0b28","2026-04-06T15:38:36",{"id":134,"version":135,"summary_zh":119,"released_at":136},333493,"2.0.0b27","2026-04-06T15:37:38",{"id":138,"version":139,"summary_zh":119,"released_at":140},333494,"2.0.0b26","2026-04-06T15:36:48",{"id":142,"version":143,"summary_zh":119,"released_at":144},333495,"2.0.0b25","2026-04-06T15:35:35",{"id":146,"version":147,"summary_zh":119,"released_at":148},333496,"2.0.0b24","2026-04-06T15:34:00",{"id":150,"version":151,"summary_zh":119,"released_at":152},333497,"2.0.0b23","2026-04-06T15:32:58",{"id":154,"version":155,"summary_zh":119,"released_at":156},333498,"2.0.0b22","2026-04-06T15:31:22",{"id":158,"version":159,"summary_zh":160,"released_at":161},333499,"2.0.0b21","Release Notes:\n   - Supporting older versions of python","2026-04-06T15:29:24",{"id":163,"version":164,"summary_zh":160,"released_at":165},333500,"2.0.0b20","2026-04-06T15:24:55",{"id":167,"version":168,"summary_zh":160,"released_at":169},333501,"2.0.0b19","2026-03-02T14:21:24",{"id":171,"version":172,"summary_zh":173,"released_at":174},333502,"2.0.0b18","Release Notes:\n  - Rebuilt the terminal emulator with selectable QProcess or Pexpect backends, added the pexpect dependency, and removed bitsandbytes for a more reliable shell experience across platforms.\n  - The engagement setup dialog is now a resizable QDialog that remembers your last folder, pre-fills Chroma\u002FThreat DB paths, and seeds the Ollama URL so you can reopen projects faster.\n  - Clear screen now wipes the central display area in both the main window and terminal emulator instead of injecting a shell command, keeping the UI tidy regardless of backend state.\n  - Hardened window lifecycle management by initializing the help window attribute and ensuring the splash\u002Fprogress window closes as soon as Nebula loads.\n  - Updated Python compatibility to 3.11–3.13.9, refreshed dependencies\u002Fdocumentation, and restored the public index page.","2025-12-02T15:59:42",{"id":176,"version":177,"summary_zh":178,"released_at":179},333503,"2.0.0b17","Release Notes:\n  - Fixes for image editing (cursor, bluriness)\n  - Enabled OpenAI within search","2025-04-14T12:03:06",{"id":181,"version":182,"summary_zh":183,"released_at":184},333504,"2.0.0b16","Release Notes:\n  - Added support for open ai\n  - Toggle between terminal and AI agent mode with a button now, you no longer have to use \"!\". You can still use \"!\" but it wil be deprecated soon\n  - Set the upper bounds of python requirements to python 3.13","2025-04-12T20:10:56",{"id":186,"version":187,"summary_zh":188,"released_at":189},333505,"2.0.0b15","Release Notes:\n  - Added an agent for tool invocation\n  - Fixed a bug where the notes window was not popping out","2025-04-11T20:39:14",{"id":191,"version":192,"summary_zh":193,"released_at":194},333506,"2.0.0b14","Release Notes:\n  - Ollama is now the sole source of LLM inference\n  - Added a thread feed that summarizes penetration testing efforts every five minutes\n  - Minor style changes\n  - Added docker image for quick setup","2025-04-10T22:13:25",{"id":196,"version":197,"summary_zh":193,"released_at":198},333507,"2.0.0b13","2025-04-10T22:07:30",{"id":200,"version":201,"summary_zh":193,"released_at":202},333508,"2.0.0b12","2025-04-10T21:59:43",[204,213,221,229,239,247],{"id":205,"name":206,"github_repo":207,"description_zh":208,"stars":209,"difficulty_score":44,"last_commit_at":210,"category_tags":211,"status":76},4358,"openclaw","openclaw\u002Fopenclaw","OpenClaw 是一款专为个人打造的本地化 AI 助手，旨在让你在自己的设备上拥有完全可控的智能伙伴。它打破了传统 AI 助手局限于特定网页或应用的束缚，能够直接接入你日常使用的各类通讯渠道，包括微信、WhatsApp、Telegram、Discord、iMessage 等数十种平台。无论你在哪个聊天软件中发送消息，OpenClaw 都能即时响应，甚至支持在 macOS、iOS 和 Android 设备上进行语音交互，并提供实时的画布渲染功能供你操控。\n\n这款工具主要解决了用户对数据隐私、响应速度以及“始终在线”体验的需求。通过将 AI 部署在本地，用户无需依赖云端服务即可享受快速、私密的智能辅助，真正实现了“你的数据，你做主”。其独特的技术亮点在于强大的网关架构，将控制平面与核心助手分离，确保跨平台通信的流畅性与扩展性。\n\nOpenClaw 非常适合希望构建个性化工作流的技术爱好者、开发者，以及注重隐私保护且不愿被单一生态绑定的普通用户。只要具备基础的终端操作能力（支持 macOS、Linux 及 Windows WSL2），即可通过简单的命令行引导完成部署。如果你渴望拥有一个懂你",349277,"2026-04-06T06:32:30",[57,55,56,212],"数据工具",{"id":214,"name":215,"github_repo":216,"description_zh":217,"stars":218,"difficulty_score":44,"last_commit_at":219,"category_tags":220,"status":76},3808,"stable-diffusion-webui","AUTOMATIC1111\u002Fstable-diffusion-webui","stable-diffusion-webui 是一个基于 Gradio 构建的网页版操作界面，旨在让用户能够轻松地在本地运行和使用强大的 Stable Diffusion 图像生成模型。它解决了原始模型依赖命令行、操作门槛高且功能分散的痛点，将复杂的 AI 绘图流程整合进一个直观易用的图形化平台。\n\n无论是希望快速上手的普通创作者、需要精细控制画面细节的设计师，还是想要深入探索模型潜力的开发者与研究人员，都能从中获益。其核心亮点在于极高的功能丰富度：不仅支持文生图、图生图、局部重绘（Inpainting）和外绘（Outpainting）等基础模式，还独创了注意力机制调整、提示词矩阵、负向提示词以及“高清修复”等高级功能。此外，它内置了 GFPGAN 和 CodeFormer 等人脸修复工具，支持多种神经网络放大算法，并允许用户通过插件系统无限扩展能力。即使是显存有限的设备，stable-diffusion-webui 也提供了相应的优化选项，让高质量的 AI 艺术创作变得触手可及。",162132,"2026-04-05T11:01:52",[55,56,57],{"id":222,"name":223,"github_repo":224,"description_zh":225,"stars":226,"difficulty_score":75,"last_commit_at":227,"category_tags":228,"status":76},1381,"everything-claude-code","affaan-m\u002Feverything-claude-code","everything-claude-code 是一套专为 AI 编程助手（如 Claude Code、Codex、Cursor 等）打造的高性能优化系统。它不仅仅是一组配置文件，而是一个经过长期实战打磨的完整框架，旨在解决 AI 代理在实际开发中面临的效率低下、记忆丢失、安全隐患及缺乏持续学习能力等核心痛点。\n\n通过引入技能模块化、直觉增强、记忆持久化机制以及内置的安全扫描功能，everything-claude-code 能显著提升 AI 在复杂任务中的表现，帮助开发者构建更稳定、更智能的生产级 AI 代理。其独特的“研究优先”开发理念和针对 Token 消耗的优化策略，使得模型响应更快、成本更低，同时有效防御潜在的攻击向量。\n\n这套工具特别适合软件开发者、AI 研究人员以及希望深度定制 AI 工作流的技术团队使用。无论您是在构建大型代码库，还是需要 AI 协助进行安全审计与自动化测试，everything-claude-code 都能提供强大的底层支持。作为一个曾荣获 Anthropic 黑客大奖的开源项目，它融合了多语言支持与丰富的实战钩子（hooks），让 AI 真正成长为懂上",160015,"2026-04-18T11:30:52",[55,57,58],{"id":230,"name":231,"github_repo":232,"description_zh":233,"stars":234,"difficulty_score":235,"last_commit_at":236,"category_tags":237,"status":76},8272,"opencode","anomalyco\u002Fopencode","OpenCode 是一款开源的 AI 编程助手（Coding Agent），旨在像一位智能搭档一样融入您的开发流程。它不仅仅是一个代码补全插件，而是一个能够理解项目上下文、自主规划任务并执行复杂编码操作的智能体。无论是生成全新功能、重构现有代码，还是排查难以定位的 Bug，OpenCode 都能通过自然语言交互高效完成，显著减少开发者在重复性劳动和上下文切换上的时间消耗。\n\n这款工具专为软件开发者、工程师及技术研究人员设计，特别适合希望利用大模型能力来提升编码效率、加速原型开发或处理遗留代码维护的专业人群。其核心亮点在于完全开源的架构，这意味着用户可以审查代码逻辑、自定义行为策略，甚至私有化部署以保障数据安全，彻底打破了传统闭源 AI 助手的“黑盒”限制。\n\n在技术体验上，OpenCode 提供了灵活的终端界面（Terminal UI）和正在测试中的桌面应用程序，支持 macOS、Windows 及 Linux 全平台。它兼容多种包管理工具，安装便捷，并能无缝集成到现有的开发环境中。无论您是追求极致控制权的资深极客，还是渴望提升产出的独立开发者，OpenCode 都提供了一个透明、可信",144296,1,"2026-04-16T14:50:03",[57,238],"插件",{"id":240,"name":241,"github_repo":242,"description_zh":243,"stars":244,"difficulty_score":75,"last_commit_at":245,"category_tags":246,"status":76},2271,"ComfyUI","Comfy-Org\u002FComfyUI","ComfyUI 是一款功能强大且高度模块化的视觉 AI 引擎，专为设计和执行复杂的 Stable Diffusion 图像生成流程而打造。它摒弃了传统的代码编写模式，采用直观的节点式流程图界面，让用户通过连接不同的功能模块即可构建个性化的生成管线。\n\n这一设计巧妙解决了高级 AI 绘图工作流配置复杂、灵活性不足的痛点。用户无需具备编程背景，也能自由组合模型、调整参数并实时预览效果，轻松实现从基础文生图到多步骤高清修复等各类复杂任务。ComfyUI 拥有极佳的兼容性，不仅支持 Windows、macOS 和 Linux 全平台，还广泛适配 NVIDIA、AMD、Intel 及苹果 Silicon 等多种硬件架构，并率先支持 SDXL、Flux、SD3 等前沿模型。\n\n无论是希望深入探索算法潜力的研究人员和开发者，还是追求极致创作自由度的设计师与资深 AI 绘画爱好者，ComfyUI 都能提供强大的支持。其独特的模块化架构允许社区不断扩展新功能，使其成为当前最灵活、生态最丰富的开源扩散模型工具之一，帮助用户将创意高效转化为现实。",109154,"2026-04-18T11:18:24",[55,56,57],{"id":248,"name":249,"github_repo":250,"description_zh":251,"stars":252,"difficulty_score":75,"last_commit_at":253,"category_tags":254,"status":76},6121,"gemini-cli","google-gemini\u002Fgemini-cli","gemini-cli 是一款由谷歌推出的开源 AI 命令行工具，它将强大的 Gemini 大模型能力直接集成到用户的终端环境中。对于习惯在命令行工作的开发者而言，它提供了一条从输入提示词到获取模型响应的最短路径，无需切换窗口即可享受智能辅助。\n\n这款工具主要解决了开发过程中频繁上下文切换的痛点，让用户能在熟悉的终端界面内直接完成代码理解、生成、调试以及自动化运维任务。无论是查询大型代码库、根据草图生成应用，还是执行复杂的 Git 操作，gemini-cli 都能通过自然语言指令高效处理。\n\n它特别适合广大软件工程师、DevOps 人员及技术研究人员使用。其核心亮点包括支持高达 100 万 token 的超长上下文窗口，具备出色的逻辑推理能力；内置 Google 搜索、文件操作及 Shell 命令执行等实用工具；更独特的是，它支持 MCP（模型上下文协议），允许用户灵活扩展自定义集成，连接如图像生成等外部能力。此外，个人谷歌账号即可享受免费的额度支持，且项目基于 Apache 2.0 协议完全开源，是提升终端工作效率的理想助手。",100752,"2026-04-10T01:20:03",[238,57,56,55]]