AI Skill Hub 推荐使用:开源协同工具 是一款优质的MCP工具。AI 综合评分 7.5 分,在同类工具中表现稳健。如果你正在寻找可靠的MCP工具解决方案,这是一个值得深入了解的选择。
开源协同工具 是一款遵循 MCP(Model Context Protocol)标准协议的 AI 工具扩展。通过 MCP 协议,它可以让 Claude、Cursor 等主流 AI 客户端直接访问和操作外部工具、数据源和服务,实现 AI 能力的无缝扩展。无论是文件操作、数据库查询还是 API 调用,都可以通过自然语言在 AI 对话中直接触发,极大提升生产效率。
开源协同工具 是一款遵循 MCP(Model Context Protocol)标准协议的 AI 工具扩展。通过 MCP 协议,它可以让 Claude、Cursor 等主流 AI 客户端直接访问和操作外部工具、数据源和服务,实现 AI 能力的无缝扩展。无论是文件操作、数据库查询还是 API 调用,都可以通过自然语言在 AI 对话中直接触发,极大提升生产效率。
# 方式一:通过 Claude Code CLI 一键安装
claude skill install https://github.com/joe-broadhead/open-cowork
# 方式二:手动配置 claude_desktop_config.json
{
"mcpServers": {
"------": {
"command": "npx",
"args": ["-y", "open-cowork"]
}
}
}
# 配置文件位置
# macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
# Windows: %APPDATA%/Claude/claude_desktop_config.json
# 安装后在 Claude 对话中直接使用 # 示例: 用户: 请帮我用 开源协同工具 执行以下任务... Claude: [自动调用 开源协同工具 MCP 工具处理请求] # 查看可用工具列表 # 在 Claude 中输入:"列出所有可用的 MCP 工具"
// claude_desktop_config.json 配置示例
{
"mcpServers": {
"______": {
"command": "npx",
"args": ["-y", "open-cowork"],
"env": {
// "API_KEY": "your-api-key-here"
}
}
}
}
// 保存后重启 Claude Desktop 生效
</div> <pre> __ __ / \ _ __ / / ____ _/ / / / / / __ \/ \/ \ / / / \ | /| / / \/ / /// / // / // / / / / / / // // / |/ |/ / // / / / , < \__/ ._/\___// // \__/\/|/|/\__// //|| // The desktop workspace for agentic work </pre>
Open Cowork is a polished Electron workspace for OpenCode — built for sessions, agents, tools, skills, workflows, artifacts, and branded downstream distributions.
Public preview: thev0.xline is intentionally unsigned while Apple Developer validation is pending. Expect rapid iteration beforev1.0.0; signed macOS builds are planned before broad distribution.
It brings the power of OpenCode into a desktop experience that developers love and less technical teams can actually use.
Docs • Getting Started • Workflows • Configuration • Open Cowork Cloud • Downstream • Roadmap
</div>
---
- Desktop-native OpenCode workspace Chat, sessions, approvals, tools, and sub-agents in one focused app.
- Agent orchestration that feels visible See delegation, tool calls, outputs, and review points directly in the transcript.
- Tools, skills, and agents in one place Manage built-ins and user-added MCPs, OpenCode skills, and custom agents from the UI.
- Project and sandbox workflows Use project threads for real filesystem work, or sandbox threads for private Cowork-managed artifacts.
- Thread-native workflows Create repeatable work by talking to Workflow Designer, then run it manually, on a schedule, or from a webhook.
- Artifact-first experience Keep generated files, outputs, and workspace artifacts organized instead of buried in chat.
- Downstream-ready distribution Configure branding, providers, defaults, bundled tools, bundled skills, docs, and release workflows.
- Provider-neutral cloud shape Run Open Cowork Cloud as all-in-one, web, worker, or scheduler roles on Docker, Helm/Kubernetes, or managed cloud infrastructure.
- Production-grade gates CI, CodeQL, smoke tests, docs, audits, checksums, SBOMs, and provenance support.
@agent.>=22.12>=10>=3.11 for docs buildspnpm install
Release artifacts are published on GitHub Releases when a tagged build completes. Signed macOS builds can check and install updates from Settings; downstream builders can configure private update release sources such as generic HTTPS feeds or Google Cloud Storage without exposing release credentials to the renderer.
Important Thev0.xpublic preview is intentionally unsigned while Apple Developer validation is pending. The release workflow can publish unsignedv0.xartifacts only when the explicit preview override is enabled; macOS will warn on first launch in that mode. To open the preview on macOS: right-click Open Cowork.app, choose Open, then choose Open again in the Gatekeeper dialog. See Apple's Gatekeeper guidance for details, or build locally.
```bash node -v
pnpm --dir apps/desktop dist:ci:mac
pnpm --dir apps/desktop dist:ci:linux
python -m pip install -r docs/requirements.txt
mkdocs build --strict
1. Download a release for your platform, or run from source. 2. Launch Open Cowork. 3. Complete first-run setup by choosing a provider, model, and whether the managed runtime can reuse standard developer config such as Git, SSH, cloud, Docker, and Kubernetes settings. 4. Connect a provider: OpenRouter API key, or OpenAI/Codex via ChatGPT Plus/Pro or API key. 5. Start a thread: - Project thread for real filesystem work in a chosen directory. - Sandbox thread for private Cowork-managed workspaces and artifacts. 6. Use @agent in the composer to invoke a sub-agent directly, or let the primary orchestrator delegate. 7. Use Workflows for repeatable work that should be saved from a Workflow Designer setup thread and run manually, on a schedule, or from a webhook.
| Home | Chat | Agents |
|---|---|---|
|  |  |  |
| Composer-first landing surface with @-agent pills. | Sub-agent delegation through @-mentions in chat. | Built-in + custom agents in one composable grid. |
| Tools & Skills | Workflows | Settings |
|---|---|---|
|  |  |  |
| Tools, skills, and MCPs with per-source visibility. | Repeatable work created from setup threads, then run manually, on schedules, or from webhooks. | Provider, model, permission, and storage controls. |
Screenshots are regenerated bypnpm screenshots— seedocs/assets/README.mdfor capture guidelines.
开源MCP工具,提供桌面工作空间,具有较高的潜力
AI Skill Hub 为第三方内容聚合平台,本页面信息基于公开数据整理,不对工具功能和质量作任何法律背书。
建议在沙箱或测试环境中充分验证后,再部署至生产环境,并做好必要的安全评估。
✅ MIT 协议 — 最宽松的开源协议之一,可自由商用、修改、分发,仅需保留版权声明。
总体来看,开源协同工具 是一款质量良好的MCP工具,在同类工具中具备一定竞争力。AI Skill Hub 将持续追踪其更新动态,建议收藏备用,结合自身场景选择合适时机引入使用。
| 原始名称 | open-cowork |
| 原始描述 | 开源MCP工具:Open Cowork is a desktop workspace for OpenCode.。⭐6 · TypeScript |
| Topics | ai-agentsautomationsdesktop-appelectronmcptypescript |
| GitHub | https://github.com/joe-broadhead/open-cowork |
| License | MIT |
| 语言 | TypeScript |
收录时间:2026-05-30 · 更新时间:2026-05-30 · License:MIT · AI Skill Hub 不对第三方内容的准确性作法律背书。
选择 Agent 类型,复制安装指令后粘贴到对应客户端