# Add to your Claude Code skills
git clone https://github.com/hAcKlyc/MyAgentsLast scanned: 5/8/2026
{
"issues": [
{
"type": "npm-audit",
"message": "@anthropic-ai/claude-agent-sdk: Vulnerability found",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "@anthropic-ai/sdk: Claude SDK for TypeScript has Insecure Default File Permissions in Local Filesystem Memory Tool",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "@chevrotain/cst-dts-gen: Vulnerability found",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "@chevrotain/gast: Vulnerability found",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "@mermaid-js/parser: Vulnerability found",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "ajv: ajv has ReDoS when using `$data` option",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "brace-expansion: brace-expansion: Zero-step sequence causes process hang and memory exhaustion",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "chevrotain: Vulnerability found",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "dompurify: DOMPurify contains a Cross-site Scripting vulnerability",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "express-rate-limit: Vulnerability found",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "flatted: flatted vulnerable to unbounded recursion DoS in parse() revive phase",
"severity": "high"
},
{
"type": "npm-audit",
"message": "hono: Hono: bodyLimit() can be bypassed for chunked / unknown-length requests",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "ip-address: ip-address has XSS in Address6 HTML-emitting methods",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "langium: Vulnerability found",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "lodash-es: Lodash has Prototype Pollution Vulnerability in `_.unset` and `_.omit` functions",
"severity": "high"
},
{
"type": "npm-audit",
"message": "mdast-util-to-hast: mdast-util-to-hast has unsanitized class attribute",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "mermaid: Vulnerability found",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "minimatch: minimatch has a ReDoS via repeated wildcards with non-matching literal in pattern",
"severity": "high"
},
{
"type": "npm-audit",
"message": "monaco-editor: Vulnerability found",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "picomatch: Picomatch: Method Injection in POSIX Character Classes causes incorrect Glob Matching",
"severity": "high"
},
{
"type": "npm-audit",
"message": "postcss: PostCSS has XSS via Unescaped </style> in its CSS Stringify Output",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "rollup: Rollup 4 has Arbitrary File Write via Path Traversal",
"severity": "high"
},
{
"type": "npm-audit",
"message": "uuid: uuid: Missing buffer bounds check in v3/v5/v6 when buf is provided",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "vite: Vite Vulnerable to Path Traversal in Optimized Deps `.map` Handling",
"severity": "high"
}
],
"status": "WARNING",
"scannedAt": "2026-05-08T05:57:28.458Z",
"semgrepRan": false,
"npmAuditRan": true,
"pipAuditRan": true
}MyAgents is an open-source ai agents skill for AI coding assistants such as Claude Code, Codex CLI, and ChatGPT, built by hAcKlyc. MyAgents - 优雅、易用的 Agent 桌面端 ,一站式 Agent 工作台与任务中心. It has 880 GitHub stars.
MyAgents returned warnings in SkillsLLM's automated security scan. It has no critical vulnerabilities, but review the flagged issues in the Security Report section before adding it to your workflow.
Clone the repository with "git clone https://github.com/hAcKlyc/MyAgents" and add it to your Claude Code skills directory (see the Installation section above).
MyAgents is primarily written in TypeScript. It is open-source under hAcKlyc on GitHub, so you can review or fork the full source.
Yes. SkillsLLM lists many other AI Agents skills you can browse and compare side by side. Open the AI Agents category from the badge at the top of this page, or use the Related Skills and comparison links further down to weigh MyAgents against similar tools.
No comments yet. Be the first to share your thoughts!
Requires a passing catalog security scan. Resolve the flagged issues and resubmit to enable featuring.
⚠️ Third-Party Software Notice
This skill is third-party open-source software developed and hosted independently on GitHub. SkillsLLM is an informational directory and does not control or maintain the underlying repository.
Any security checks, ratings, or warnings displayed by SkillsLLM are automated and limited in scope. They do not constitute a security certification or guarantee that the software is safe, error-free, or free from malicious code, vulnerabilities, compromised dependencies, or prompt-injection risks.
Review the source code, permissions, dependencies, and configuration before installing or running any third-party skill. Use is at your own risk. To the maximum extent permitted by applicable law, SkillsLLM is not liable for losses arising from third-party software.
活在你的电脑里,真正能干活的个人 Agent
中文 · English · 官网 · 下载 · 架构文档 · 贡献指南

MyAgents 是一款开源桌面端个人 Agent 工作台。它不是另一个聊天窗口,而是把对话、工作区、文件、工具、模型、任务和长期记忆放进同一个桌面系统里,让 AI 真正进入你的日常工作流。
你可以把它理解成三件事的组合:
最终目标很简单:让人把注意力放在判断、品味和验收上,把上下文整理、工具调用、长程执行和重复工作交给 Agent。
每个 Agent 都绑定真实工作区。对话不再漂浮在一个孤立输入框里,AI 可以围绕文件、命令、技能、工具和历史上下文持续工作。

在一个会话里,MyAgents 同时提供:
@ 引用文件、/ 调用技能、MCP 工具调用、定时任务入口和模型选择。MyAgents 内置「想法 + 任务中心」。你可以先把零散念头记下来,再和 AI 讨论、对齐目标、沉淀成可执行任务。长期任务不需要留在聊天记录里翻找,而是进入可追踪的状态机。

任务中心支持:
MyAgents 不把用户锁死在单一模型或单一供应商里。你可以使用 Anthropic 订阅或 API,也可以配置 DeepSeek、Moonshot、智谱、MiniMax、Google Gemini、火山方舟、硅基流动、ZenMux、OpenRouter、小米 MiMo、阿里云百炼等供应商。实际模型列表以应用内「模型供应商」页为准。

除了模型,MyAgents 还支持:
桌面 AI 不应该只在你打开主应用时才存在。MyAgents 提供小助理、桌面宠物/浮窗、IM Bot 和定时任务,让 Agent 能在不同入口里承接同一个工作上下文。

你可以在主窗口里做长对话,也可以在桌面浮窗里快速发问;可以让 Agent 在 IM 里处理消息,也可以让它按计划自动执行任务。MyAgents 关注的不是「多一个聊天入口」,而是让 AI 能进入真实的工作节奏。
过去很多 AI 产品把「对话」当成唯一形态。对话很自然,但它不适合承载长期工作:上下文会散、任务会丢、结果难复盘,最后用户又回到手工整理。
我更希望 MyAgents 把 Agent 看成一个持续工作的系统。聊天只是入口,真正重要的是工作区、文件、工具、任务、状态和记忆。一个 Agent 应该能知道自己在哪个项目里、正在做什么、上次做到哪里、下一步该验证什么。
AI 最有价值的地方不是替人多生成几段文字,而是把混乱信息整理成可判断、可执行、可沉淀的东西。
所以 MyAgents 里有想法和任务中心。想法用于收集不成熟的判断,任务用于承载已经确认的目标。中间的讨论、计划、执行、验证都可以交给 Agent,但最后的方向感和验收标准仍然留给人。
一个桌面 Agent 不应该只复制网页聊天体验。它应该能接触本地文件、终端、浏览器、通知、定时任务、IM 和系统环境,同时保持边界清晰、权限可控、数据本地优先。
MyAgents 的很多设计都来自这个判断:本地工作区是一等公民,Sidecar 按 Session 隔离,所有文件能力走 Tauri/Rust,AI Runtime 可以切换,模型供应商可以替换,工具和 Skills 可以扩展。
Agent 产品不可能预设所有人的工作流。开发者、创作者、研究者、产品经理、教育工作者和行业专家需要的能力都不一样。与其做一个「什么都内置但什么都固定」的应用,不如提供一个稳定的底座,让用户把自己的工具、模型、技能和自动化流程接进来。
这也是 MyAgents 坚持开源、支持 MCP、Skills、插件和多供应商的原因。
| 能力 | 说明 |
|---|---|
| 多标签 Agent | 每个 Tab 独立会话和 Sidecar,适合并行工作 |
| 工作区系统 | 文件树、预览、搜索、Git 分支、Skills 和命令统一入口 |
| 多模型供应商 | Anthropic 订阅/API、多家国内外 API、OpenRouter/ZenMux 等聚合服务 |
| 多 Agent Runtime | 内置 Claude Agent SDK,可选 Claude Code CLI / Codex CLI / Gemini CLI |
| MCP 工具 | 支持 STDIO / HTTP / SSE,内置和外部 MCP 可并存 |
| Skills | 内置技能、用户技能、工作区技能,适合沉淀固定流程 |
| 任务中心 | 想法、任务、周期调度、状态追踪和执行审计 |
| IM Bot / Channel | Telegram、钉钉、OpenClaw 插件 Channel |
| 内嵌终端 | xterm.js + portable-pty,绑定当前工作区 |
| 内嵌浏览器 | Tauri 多 Webview 子视图,方便预览链接和本地 HTML |
| 本地全文搜索 | Tantivy + jieba,检索会话历史和工作区文件 |
| 本地优先 | 会话、任务、配置和生成产物默认保存在本机 |
MyAgents 是一个桌面端 AI Agent 产品,不是单纯的前端项目。改动前建议先判断你碰到的是 UI、Rust 桌面层、Node Sidecar、Agent Runtime、MCP、任务中心还是插件桥接。
| 层级 | 技术 |
|---|---|
| 桌面框架 | Tauri v2 + Rust |
| 前端 | React 19 + TypeScript + Vite + TailwindCSS |
| 后端 Sidecar | Node.js v24 + Claude Agent SDK |
| 通信 | Rust HTTP/SSE Proxy,前端通过 Tauri invoke 代理到 Sidecar |
| Runtime | 内置 Claude Agent SDK,实验室支持 Claude Code CLI / Codex CLI / Gemini CLI |
| 工具生态 | MCP、Skills、OpenClaw Plugin Bridge、myagents CLI |
| 搜索 | Tantivy + tantivy-jieba |
| 终端 | portable-pty + xterm.js |
最终用户:
开发者:
>=22.0.0,推荐 Node.js 24。npm@11.13.0。macOS / Linux:
git clone https://github.com/hAcKlyc/MyAgents.git
cd MyAgents
./setup.sh
./start_dev.sh
Windows:
git clone https://github.com/hAcKlyc/MyAgents.git
cd MyAgents
.\setup_windows.ps1
.\build_windows.ps1
setup 脚本会先按当前 target 和已验证缓存检查原生推理构建工具,再开始下载 Node runtime、安装项目依赖或拉取 Rust crates;缺失的原生构建工具会得到安装与验证提示,不会被脚本自动安装。随后它会缓存离线文档与语音推理资源;重复运行会复用已验证的资源缓存。默认 Mino 工作区模板已经提交在 bundled-workspaces/mino/,构建和初始化不需要额外的 GitHub SSH 或模板下载。
# 启动开发环境
./start_dev.sh
# 类型检查
npm run typecheck
# Lint,包含 ESLint 和 dependency-cruiser 架构边界检查
npm run lint
# 测试分层
npm run test:classification # server 测试命名/分层 guard
npm run test:unit # 纯逻辑快池
npm run test:dom # React/jsdom 组件与 hook
npm run test:integration # CI-safe 后端集成池,无真实网络/密钥
npm test # classification + unit + dom + integration
npm run test:credentialed # 真实 Provider/SDK smoke,显式本地运行
# Debug 构建,含 DevTools
./build_dev.sh
# macOS 生产构建
./build_macos.sh
# Linux AppImage + deb 构建
./build_linux.sh
Linux 构建机需要 Tauri/WebKit 相关系统依赖,Ubuntu/Debian 可参考:
sudo apt-get install -y \
build-essential libssl-dev libgtk-3-dev libayatana-appindicator3-dev \
librsvg2-dev libwebkit2gtk-4.1-dev patchelf
src/renderer/ React 前端
src/server/ Node.js Sidecar
src/server/plugin-bridge/ OpenClaw Plugin Bridge
src/cli/ myagents CLI
src/shared/ 前后端共享类型
src-tauri/ Tauri Rust 层
bundled-agents/ 内置 Agent
bundled-skills/ 内置 Skills
specs/ 架构、设计、技术文档
/refs/:id、/attachment/* 大载荷数据面在 CORS/CSP/路径安全约束下原生 fetch。src/server/session-engine/ facade,由 selector 选择 builtin 或 external adapter;不能让 builtin SDK resume 外部会话。完整架构请读 specs/ARCHITECTURE.md。具体模块请按需阅读:
提交前至少运行:
npm run typecheck
npm run lint
npm run test:classification
npm run test:unit
后端 Session、Runtime、IO 或安全边界改动还应跑 npm run test:integration;真实 Provider / SDK 链路只在本机显式跑 npm run test:credentialed,不属于默认 CI。
如果改动涉及 Rust、Tauri 命令、Sidecar 生命周期、Runtime、MCP、任务中心或插件桥接,请先阅读对应 specs/tech_docs/ 文档,避免绕开已有架构。
提交信息遵循 Conventional Commits:
feat: add ...
fix: handle ...
docs: update ...
refactor: simplify ...
test: cover ...
chore: bump ...
MyAgents 采用 GNU Affero General Public License v3.0
(AGPL-3.0-only)。个人和企业都可以免费使用,也可以商业使用;分发修改版、通过网络向
用户提供修改版等场景需要完整履行 AGPL,包括在适用时提供对应源码。
如果你希望在不履行 AGPL 开源义务的情况下闭源修改、嵌入、OEM、分发或托管 MyAgents, 需要取得单独的商业许可证。请联系 myagents.io@gmail.com。
完整说明见 LICENSING.md,商业许可概要见 COMMERCIAL-LICENSING.md。第三方组件继续适用各自许可证,详见 THIRD_PARTY_NOTICES.md。
MyAgents is an open-source desktop workspace for personal AI Agents. It is not another chat window. It puts conversations, workspaces, files, tools, models, tasks, and long-term memory into one desktop system, so AI can become part of your real daily workflow.
You can think of it as three things in one: