jungle-grid-mcp-server
OfficialJungle Grid MCP Server
从支持 MCP 的 AI 主机(如 Claude Desktop、Cursor、Windsurf 和 MCP Inspector)运行 Jungle Grid GPU 工作负载。
该服务器通过 stdio 在本地运行,并将工具调用转发给带有您 API 密钥的 Jungle Grid REST API。
要求
Node.js 18 或更高版本
Jungle Grid API 密钥
可选:用于自托管编排器的
JUNGLE_GRID_API_URL
对于完整的提交工作流,API 密钥需要 jobs:write 权限。该范围允许对密钥账户拥有的作业进行估算、提交、轮询、取消和日志查看。list_jobs 仍然需要 jobs:read 权限。
Related MCP server: jlab-mcp
快速入门
JUNGLE_GRID_API_KEY=jg_... npx -y @jungle-grid/mcp在 Windows PowerShell 上:
$env:JUNGLE_GRID_API_KEY = "jg_..."
npx -y @jungle-grid/mcp该服务器使用 stdio,因此手动启动成功后会显示为等待 MCP 消息。如果缺少 JUNGLE_GRID_API_KEY,它将以明确的错误退出。
Claude Desktop
将其添加到 claude_desktop_config.json,然后完全重启 Claude Desktop。
{
"mcpServers": {
"junglegrid": {
"command": "npx",
"args": ["-y", "@jungle-grid/mcp"],
"env": {
"JUNGLE_GRID_API_KEY": "jg_..."
}
}
}
}Windows 配置文件路径:
%APPDATA%\Claude\claude_desktop_config.jsonmacOS 配置文件路径:
~/Library/Application Support/Claude/claude_desktop_config.jsonCursor 或 Project MCP 配置
对于已签入的项目配置,请避免提交密钥。将 API 密钥放入用于启动 Cursor 的环境中,并保持配置不包含密钥。
{
"mcpServers": {
"junglegrid": {
"command": "npx",
"args": ["-y", "@jungle-grid/mcp"]
}
}
}对于本地未提交的配置,您可以直接包含密钥:
{
"mcpServers": {
"junglegrid": {
"command": "npx",
"args": ["-y", "@jungle-grid/mcp"],
"env": {
"JUNGLE_GRID_API_KEY": "jg_..."
}
}
}
}自托管编排器
JUNGLE_GRID_API_URL 默认为 https://api.junglegrid.dev。当您的主机需要调用不同的编排器时,请覆盖此设置。
{
"mcpServers": {
"junglegrid": {
"command": "npx",
"args": ["-y", "@jungle-grid/mcp"],
"env": {
"JUNGLE_GRID_API_KEY": "jg_...",
"JUNGLE_GRID_API_URL": "https://your-orchestrator.example.com"
}
}
}
}工具
estimate_job:估算 GPU 层级、区域、持续时间和积分成本。submit_job:提交异步 GPU 工作负载,并可选择environment值。get_job:获取当前作业状态和详细信息。list_jobs:列出已认证账户的近期作业。cancel_job:取消挂起、排队或运行中的作业。get_job_logs:获取 stdout、stderr 和退出信息。stream_job_logs:流式传输实时日志,直到完成或超时。list_job_artifacts:列出为作业上传的托管工件。get_artifact_download_url:为单个托管工件创建签名的下载 URL。
实时作业模式
使用 submit_job 开始工作,使用 stream_job_logs 获取实时输出,然后在完成后使用 list_job_artifacts 检索保存的文件。
{
"command": ["python", "-c", "import os; exec(os.environ['CODE'])"],
"environment": {
"CODE": "import os, json\nos.makedirs('/workspace/artifacts', exist_ok=True)\nwith open('/workspace/artifacts/output.json','w') as f:\n json.dump({'status':'ok'}, f)"
}
}当实际的 Python 负载太长而无法舒适地放入 command 数组时,这是推荐的模式。
对于托管作业,Jungle Grid 会自动创建 /workspace/artifacts 并上传写入其中的任何常规文件。用户无需手动创建签名的上传 URL 或调用工件完成端点。
本地开发
npm install
npm run build
JUNGLE_GRID_API_KEY=jg_... node dist/index.js使用 MCP Inspector 检查服务器:
JUNGLE_GRID_API_KEY=jg_... npx @modelcontextprotocol/inspector node dist/index.js发布
在发布前验证包:
npm run build
npm pack --dry-run公开发布作用域包:
npm publish --access public故障排除
JUNGLE_GRID_API_KEY environment variable is required:将密钥添加到主机配置的env块中,或添加到启动主机的环境中。工具未显示:编辑配置后,请完全退出并重新打开 MCP 主机。
旧包版本:在配置中固定版本,例如
["@jungle-grid/mcp@0.1.0"],或清除 npx 缓存。API 调用失败:确认密钥有效且
JUNGLE_GRID_API_URL指向您打算使用的编排器。
欢迎贡献者
我们正在向对 AI 智能体、MCP、开发工具和工作负载执行感兴趣的贡献者开放 Jungle Grid MCP 服务器。
良好的切入领域:
改进文档
添加示例提示词
为 MCP 工具处理程序添加测试
添加 Docker 支持
改进 GitHub Actions
构建集成示例
从标记为 good first issue 的议题开始。
Maintenance
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Related MCP Servers
- AlicenseAqualityDmaintenanceAn MCP server that enables AI agents to interact with Modal, allowing them to deploy apps and run functions in a serverless cloud environment.73MIT
- FlicenseAqualityBmaintenanceAn MCP server that enables LLMs to execute Python code on GPU-accelerated compute nodes within SLURM-managed HPC environments. It bridges local clients to remote clusters by launching JupyterLab sessions via SLURM jobs to facilitate high-performance notebook-based computation.7
- FlicenseNot gradedqualityAmaintenanceAn MCP server for monitoring and managing multi-cluster Slurm GPU jobs, enabling AI agents to execute commands, check allocations, and explore logs across HPC clusters.1
- AlicenseNot gradedqualityBmaintenanceMCP server that enables agents to dynamically switch between multiple AI models (OpenAI, Anthropic, Google, etc.) with unified protocol-driven configuration and capability discovery.Apache 2.0
Related MCP Connectors
HiveCompute MCP Server — decentralized inference router for AI agents
MCP server connecting AI agents to non-custodial staking data across 130+ networks.
MCP server for AI agents to plan, verify, and deploy Cloudflare-native apps.
Latest Blog Posts
- Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)By Om-Shree-0709 on .mcpAgent IdentityOAuth 2.1
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/Jungle-Grid/mcp-server'
If you have feedback or need assistance with the MCP directory API, please join our Discord server