ZimaTools
Provides tools for managing Docker containers on ZimaOS over SSH, including listing, starting, stopping, restarting, viewing logs, and retrieving container information.
Includes a legacy/historical Home Assistant integration located under legacy/ha/.
Provides GPU lease management and status reporting via nvidia-smi, including exclusive GPU acquisition, release, queue listing, and client priority adjustments.
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@ZimaToolsshow me the GPU status and current lease queue"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
ZimaTools
Application native ZimaOS : serveur MCP (HTTP), interface web et arbitre GPU pour une VRAM exclusive.
Popcorn a priorite sur Ollama et les agents. Un seul lease GPU a la fois.
Repo : https://github.com/bobdivx/zimatools
Origine : zimaos-cursor-mcp v0.0.2 (historique conserve).
Stack
Monorepo pnpm workspaces
apps/mcp— TypeScript, mcp-framework0.2.x, transport HTTP Stream (/mcp) + stdio optionnelPont REST Hono (
/api/*) pour l'UIapps/web— Astro + Preact + Tailwind CSS + DaisyUIdocker-compose.ymlpret pour ZimaOS (docker.sock, NVIDIA optionnel)
Related MCP server: AX Local Operations MCP Server
Architecture
zimatools/
apps/mcp/ serveur MCP + REST + arbitre GPU
apps/web/ UI Astro/Preact/DaisyUI
docker/ Dockerfiles
docker-compose.yml
docker-compose.gpu.ymlService | Port | Role |
MCP HTTP Stream | 8765 | Cursor, DevForge, agents distants |
REST API | 8766 | UI (GPU, Docker) |
Web UI | 8080 | Tableau de bord |
Installation locale (dev)
Prerequis : Node.js 20+, pnpm 10 (corepack enable).
git clone https://github.com/bobdivx/zimatools.git
cd zimatools
cp .env.example .env
# renseigner ZIMAOS_API_BASE / ZIMAOS_API_TOKEN / SSH si besoin
pnpm install
pnpm --filter @zimatools/mcp buildDeux terminaux :
pnpm dev:mcp # MCP HTTP :8765 + REST :8766
pnpm dev:web # UI :4321 (proxy /api -> 8766)Stdio (Cursor local uniquement) :
pnpm --filter @zimatools/mcp dev:stdioInstallation sur ZimaOS
Pas de deploiement automatique dans cette version. Compose pret a copier :
Cloner (ou copier) le depot sur le NAS.
cp .env.example .envet renseigner les tokens.Sans GPU :
docker compose up -d --buildAvec NVIDIA (nvidia-container-toolkit) :
docker compose -f docker-compose.yml -f docker-compose.gpu.yml up -d --buildUI :
http://<nas>:8080MCP :
http://<nas>:8765/mcpREST :
http://<nas>:8766/health
Le socket Docker est monte (/var/run/docker.sock). Le runtime NVIDIA est optionnel.
Endpoint MCP (HTTP)
Transport : Streamable HTTP (mcp-framework http-stream), pas seulement stdio.
URL :
http://<nas-ou-localhost>:8765/mcpMethodes :
POST/GET/DELETE/OPTIONSCORS ouvert par defaut (
MCP_CORS_ORIGIN=*,MCP_HOST=0.0.0.0)
Exemple Cursor / DevForge (mcp.json) :
{
"mcpServers": {
"zimatools": {
"url": "http://zimacube.local:8765/mcp"
}
}
}Stdio local (sans NAS) :
{
"mcpServers": {
"zimatools": {
"command": "node",
"args": ["<chemin>/apps/mcp/dist/index.js", "--stdio"],
"env": {
"ZIMAOS_API_BASE": "http://zimacube.local",
"ZIMAOS_API_TOKEN": "..."
}
}
}
}REST UI (Hono, memes outils GPU) :
GET /healthGET /api/gpu/statusPOST /api/gpu/acquire{ "client": "popcorn" }POST /api/gpu/release{ "client": "popcorn" }GET /api/gpu/queuePOST /api/gpu/priority{ "client": "ollama", "priority": 50 }GET /api/docker/containers
Outils MCP
Fichiers ZimaOS (existants, via API)
read_file_from_zimaos, list / write / edit / search / mkdir / stats — voir apps/mcp/src/tools/.
Docker (existants, via SSH)
list_docker_containers_zimaos, start / stop / restart / logs / info.
GPU (v0 — file en memoire + nvidia-smi)
Outil | Role |
| nvidia-smi + lease + file. Si |
| Lease exclusif. Priorite plus haute preempte. |
| Relache et promeut la file. |
| Lease courant + file. |
| Priorite client (defaut popcorn=100, ollama=50, agents=25). |
Modele : un seul holder. Popcorn vole le GPU a Ollama / agents ; le detenteur precedent est refile.
Stub / limites v0
Queue et priorites en memoire (perdues au restart).
nvidia-smiparse CSV ; sinon GPU facticestub-gpu.Pas encore d'arret/redemarrage reel d'Ollama ou de containers GPU.
Liste Docker UI : SSH si configure, sinon placeholder.
Variables d'environnement
Voir .env.example.
Variable | Defaut | Role |
|
|
|
|
| MCP Streamable HTTP |
|
| Bind NAS |
|
| Chemin MCP |
|
| REST Hono |
| hostname:8766 | URL API vue par le navigateur |
| — | Outils fichiers |
| — | Outils Docker |
Developpement
pnpm build # mcp + web
pnpm --filter @zimatools/web buildL'integration Home Assistant historique est dans legacy/ha/.
Licence
A definir. Issues : https://github.com/bobdivx/zimatools/issues
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections. Dates show when Glama detected each change.
No tool schema history has been recorded yet.
This server cannot be installed
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 Connectors
Protocol-native energy infrastructure orchestration for AI data centers. Provides 46 MCP tools across 8 grid protocols (IEC-61850, DNP3, Modbus, OCPP, OpenADR, IEEE 2030.5, IEC 60870-5-104, ICCP) with 5 core API primitives: connect, dispatch, settle, comply, and intel. Enables AI agents to programmatically interact with substations, grid interfaces, and energy assets for real-time workload-grid coordination.
Paid remote MCP for persistent AI agent memory, analytics, checkout, and search-readiness.
OCR, transcription, file extraction, and image generation for AI agents via MCP.
CPU telemetry & benchmarking, SHA-256 hashing, AES-256-GCM storage, AI video/image via Nova.
Related MCP Servers
- FlicenseNot gradedqualityDmaintenanceEnables offline AI agent automation with embedded local LLM (Qwen 2.5), sandboxed file operations through AgentFS, and dynamic skill loading. Exposes capabilities via MCP with tri-state safety guards for private, air-gapped environments without network connectivity or API costs.-
- AlicenseAqualityDmaintenanceA comprehensive MCP server that enables AI models to perform local file operations, command execution, and task management across multiple platforms. It features advanced capabilities like row-level file editing, directory searching, and system monitoring with built-in security filters.1313Mulan Permissive Software , Version 2
- FlicenseNot gradedqualityDmaintenanceGives AI agents real-time access to system metrics, process management, and container orchestration.-
- AlicenseNot gradedqualityAmaintenanceEnables AI agents to manage system updates, application installations, and remote host orchestration through MCP tools.3MIT
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/bobdivx/zimatools'
If you have feedback or need assistance with the MCP directory API, please join our Discord server