E2E Networks Cloud & TIR MCP Server
Connect AI assistants to E2E Networks Cloud and TIR for authenticated, tool-based cloud and AI/ML infrastructure management.
Provision, inspect, list, power on/off, reboot, lock/unlock, rename, reinstall, and delete compute/GPU nodes; view plans, OS images, and node health.
Manage TIR AI/ML resources: list GPU SKUs, launch/control Jupyter AI Labs, monitor model endpoints, datasets, and Slurm/distributed training clusters.
Manage storage: block volumes (create, attach, detach, delete), EOS S3 buckets, and SFS shared file systems.
Manage networking/security: VPCs, reserved IPs, security groups, and load balancers.
Manage DBaaS databases (PostgreSQL, MySQL, MariaDB, Kafka, etc.) and Kubernetes clusters/node pools.
Inspect projects, billing summary, test connectivity, configure credentials/profiles, and run via npm, Docker, or HTTP/SSE daemon mode.
Use e2e_raw_request to call any E2E REST API endpoint with automatic auth and region/project handling.
Apply guardrails: read-only mode, tool allow/deny lists, confirmation for destructive/state-changing actions, and MCP annotations.
Provides tools for launching and managing Jupyter AI Labs on the TIR AI/ML Platform, including listing, creating, and controlling notebook instances.
Provides tools for inspecting and managing managed Kubernetes clusters and worker node pools on E2E Cloud.
Enables provisioning and managing managed MariaDB database clusters as part of E2E Cloud DBaaS.
Enables provisioning and managing managed MySQL database clusters as part of E2E Cloud DBaaS.
Provides tools for listing and managing NVIDIA GPU instances and hardware SKUs (H100, A100, L40S, L4) on E2E Cloud and TIR.
Enables provisioning and managing managed PostgreSQL database clusters as part of E2E Cloud DBaaS.
Click on "Deploy 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., "@E2E Networks Cloud & TIR MCP Server launch an A100 GPU node in DEL-1 and give me its public IP"
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.
E2E Networks Cloud & TIR MCP Server
The community Model Context Protocol (MCP) server for E2E Networks Cloud and the TIR AI/ML Platform.
Connect Claude Desktop, Cursor, Google Antigravity, and Codex directly to your E2E Cloud infrastructure to provision GPUs, manage nodes, control databases, orchestrate Kubernetes, and run AI workloads.
⚡ Quickstart (npm install)
Step 1: Install globally via npm
npm install -g e2e-mcpStep 2: Configure your E2E credentials
Run the interactive setup wizard (like aws configure):
e2e-mcp configureThis prompts for:
E2E API Key (from MyAccount → Security / API Tokens)
E2E Auth Token (Bearer token)
Default Location (e.g.
DEL-1orNCR-1)Saved securely to
~/.e2e/credentials(chmod 0600).
Step 3: Auto-register with your AI Assistant
Automatically register the MCP server in Claude Desktop, Cursor, Google Antigravity, and Codex:
e2e-mcp installRestart your AI assistant, and you're ready!
Related MCP server: Cloud Pilot MCP
💻 Manual Setup in AI Assistants
If you prefer to configure your assistant's JSON file manually:
1. Claude Desktop
Add to your config (~/Library/Application Support/Claude/claude_desktop_config.json on macOS or %APPDATA%\Claude\claude_desktop_config.json on Windows):
{
"mcpServers": {
"e2e-cloud": {
"command": "e2e-mcp"
}
}
}If you didn't rune2e-mcp configure, you can pass credentials directly in the env block:
"env": {
"E2E_API_KEY": "YOUR_API_KEY",
"E2E_AUTH_TOKEN": "YOUR_AUTH_TOKEN",
"E2E_PROJECT_ID": "YOUR_PROJECT_ID",
"E2E_LOCATION": "DEL-1"
}2. Cursor
Add to .cursor/mcp.json or Cursor Settings → MCP:
{
"mcpServers": {
"e2e-cloud": {
"command": "e2e-mcp"
}
}
}3. Google Antigravity
Add to ~/.gemini/config/mcp_config.json:
{
"mcpServers": {
"e2e-cloud": {
"command": "e2e-mcp"
}
}
}🛠️ Available MCP Tools
Once connected, your AI assistant gains access to 40+ specialized cloud tools:
🖥️ Compute & NVIDIA GPUs
e2e_list_nodes— List all running VMs and GPU instances with IPs, plans, and regions.e2e_get_node— Get full technical specifications, network interfaces, and disk info for a node.e2e_create_node— Launch new compute or GPU instances on-demand.e2e_node_action— Control node lifecycle:power_on,power_off,reboot,lock,unlock.e2e_delete_node— Terminate and delete an instance.e2e_list_plans/e2e_list_os_images— Inspect hardware plans, pricing, and operating systems.e2e_get_node_health— Retrieve live CPU, memory, and disk health metrics.
🤖 TIR AI/ML Cloud
e2e_tir_list_gpu_skus— List available NVIDIA GPU hardware (H100 SXM5, A100, L40S, L4).e2e_tir_list_notebooks/e2e_tir_create_notebook— Launch and manage Jupyter AI Labs.e2e_tir_notebook_action— Start, stop, or reboot AI Lab notebooks.e2e_tir_list_model_endpoints— Monitor deployed LLM inference endpoints.e2e_tir_list_datasets— Manage training and fine-tuning datasets.e2e_tir_list_training_clusters— Manage Slurm & distributed training clusters.
💾 Storage & Buckets
e2e_list_volumes/e2e_create_volume— Manage block storage volumes (SSD/NVMe).e2e_attach_volume/e2e_detach_volume— Attach or detach block volumes to nodes.e2e_list_buckets/e2e_create_bucket— S3-compatible EOS Object Storage buckets.e2e_list_sfs— Shared File System (SFS/NFS) storage.
🌐 Networking & Security
e2e_list_vpcs/e2e_create_vpc— Virtual Private Cloud networks and subnets.e2e_list_reserved_ips/e2e_action_reserved_ip— Manage static public IP addresses.e2e_list_security_groups— Inspect firewalls and security rules.e2e_list_load_balancers— Application and Network Load Balancers (ALB / NLB).
🗄️ Managed Databases (DBaaS)
e2e_list_databases/e2e_create_database— Provision managed PostgreSQL, MySQL, MariaDB, Kafka.e2e_get_database/e2e_database_action— Cluster health, topologies, failover, start/stop.
☸️ Managed Kubernetes
e2e_list_k8s_clusters/e2e_get_k8s_cluster— Inspect Kubernetes clusters and API endpoints.e2e_list_k8s_node_pools— Manage cluster worker node pools.
🌐 Universal REST Tool (e2e_raw_request)
e2e_raw_request— Allows Claude to call any of E2E's 450+ REST endpoints with automatic authentication, region normalization, and error handling.
🔒 Read-only mode & tool annotations
Set E2E_READ_ONLY=true in the server's environment to expose only the list/get tools. Create, delete, attach, detach and action tools (and e2e_configure_credentials) are not registered, and e2e_raw_request rejects any method other than GET.
Limit which tools exist. E2E_ENABLED_TOOLS (allowlist) and E2E_DISABLED_TOOLS (denylist) take comma-separated tool names or groups: compute, storage, network, database, kubernetes, tir, platform, raw. Example: E2E_ENABLED_TOOLS=compute,network or E2E_DISABLED_TOOLS=raw,e2e_delete_node. The denylist wins when both match.
Confirm step. Destructive and state-changing tools (and non-GET e2e_raw_request calls) take a confirm argument and refuse to run unless it is true, returning a message telling the assistant to get the user's explicit approval first. This is a guardrail against accidental or prompt-injected calls, not a hard boundary, since the model supplies the argument. For a hard boundary use read-only mode or the allowlist. Set E2E_REQUIRE_CONFIRM=false for unattended automation.
Every tool also carries MCP annotations (readOnlyHint, destructiveHint, idempotentHint) so clients can prompt before destructive calls. Tools not explicitly classified default to destructive.
⚙️ Optional & Advanced Setups
You can maintain separate profiles (e.g. [default], [production], [staging]):
e2e-mcp configure --profile staging
e2e-mcp install --profile stagingStored in ~/.e2e/credentials and ~/.e2e/config.
Run as a shared HTTP service with SSE and a built-in Web Dashboard:
e2e-mcp --http --port 3000Dashboard:
http://localhost:3000/SSE Endpoint:
http://localhost:3000/sseHealth Check:
http://localhost:3000/health
docker run -d -p 3000:3000 \
-e E2E_API_KEY="your_key" \
-e E2E_AUTH_TOKEN="your_token" \
ghcr.io/vingupta3/e2e-mcp:latestgit clone https://github.com/vingupta3/e2e-mcp.git
cd e2e-mcp
npm install
npm test
npm run buildSee CONTRIBUTING.md for pull request guidelines.
📄 License
Apache 2.0 © 2026 Vinay Gupta
Available Tools
46 toolse2e_action_reserved_ipC
Attach, detach, or live-reserve a static public IP to/from a compute node or load balancer.
Inputs: type (required), vm_id (required), ip_address (required), location, +1 more.
e2e_attach_security_groupC
Attach a security group to a compute node.
Inputs: vm_id (required), security_group_id (required), location, project_id.
e2e_attach_volumeC
Attach a block storage volume to a compute node (VM).
Inputs: vm_id (required), block_id (required), location, project_id.
e2e_configure_credentialsA
Configure E2E Networks API credentials in ~/.e2e/credentials and ~/.e2e/config. Allows any AI assistant (Google Antigravity, Claude, Cursor, Codex) to set up or update API credentials directly. location and project_id are optional and should be omitted to allow searching across all projects and locations.
Inputs: api_key (required), auth_token (required), profile, location, +1 more.
e2e_create_bucketC
Create an EOS (E2E Object Storage) bucket.
Inputs: bucket_name (required), location, project_id, is_lock_enabled, +2 more.
e2e_create_databaseC
Provision a managed database cluster (MySQL, PostgreSQL, etc.).
Inputs: name (required), software_id (required), template_id (required), vpcs, +3 more.
e2e_create_nodeC
Launch and provision a new compute node or GPU instance on E2E Cloud.
Inputs: name (required), plan (required), image (required), disk, +10 more.
e2e_create_volumeC
Create an independent block storage volume on E2E Cloud.
Inputs: name (required), size (required), iops, location, +1 more.
e2e_create_vpcC
Create a new Virtual Private Cloud (VPC) network.
Inputs: vpc_name (required), ipv4, location, is_e2e_vpc, +1 more.
e2e_database_actionC
Execute a control action (start, stop, restart) on a managed database cluster.
Inputs: action (required), cluster_id (required), location, project_id.
e2e_delete_bucketA
Delete an EOS object storage bucket. The bucket must be empty before deletion.
Inputs: bucket_name (required), location, project_id.
e2e_delete_nodeA
Terminate and delete a compute node from E2E Cloud. Warning: This permanently removes the node and its local disk.
Inputs: node_id (required), location, project_id.
e2e_delete_volumeA
Permanently delete an unattached block storage volume.
Inputs: block_id (required), location, project_id.
e2e_delete_vpcC
Delete a VPC network by its network ID.
Inputs: vpc_network_id (required), location, project_id.
e2e_detach_security_groupC
Detach a security group from a compute node.
Inputs: vm_id (required), security_group_id (required), location, project_id.
e2e_detach_volumeC
Detach a block storage volume from a compute node.
Inputs: vm_id (required), block_id (required), location, project_id.
e2e_get_billing_summaryC
Get estimated monthly usage, prepaid credits balance, or billing summary.
Inputs: location, project_id.
e2e_get_databaseB
Get detailed information, topology, connection endpoints, and health of a managed database cluster.
Inputs: cluster_id (required), location, project_id.
e2e_get_k8s_clusterC
Get detailed information about a managed Kubernetes cluster by its service ID.
Inputs: service_id (required), location, project_id.
e2e_get_nodeA
Get detailed information about a specific compute node by its ID (specs, status, public/private IPs, attached disks, security groups).
Inputs: node_id (required), location, project_id.
e2e_get_node_healthB
Retrieve server health information and metrics (CPU utilization, memory usage, disk space) for a node.
Inputs: node_id (required), location, project_id.
e2e_list_bucketsB
List all EOS (E2E Object Storage) S3-compatible buckets in your account.
Inputs: location, project_id.
e2e_list_database_plansB
List available DBaaS database sizing plans, engine versions, and pricing.
Inputs: location, project_id.
e2e_list_databasesA
List all managed database clusters (DBaaS) in your project (MySQL, PostgreSQL, MariaDB, Kafka, Valkey, OpenSearch).
Inputs: location, project_id.
e2e_list_k8s_clustersB
List all managed Kubernetes clusters in your E2E Cloud project.
Inputs: location, project_id.
e2e_list_k8s_node_poolsC
List all node pools attached to a managed Kubernetes cluster.
Inputs: service_id (required), location, project_id.
e2e_list_load_balancersC
List all managed Load Balancers (Application Load Balancers and Network Load Balancers) in your project.
Inputs: location, project_id.
e2e_list_nodesB
List all compute nodes (virtual machines and GPU instances) in your E2E Cloud project, including status, IP addresses, plans, and regions.
Inputs: location, project_id.
e2e_list_os_imagesB
List available OS categories, distributions (Ubuntu, Debian, CentOS, AlmaLinux, Windows), and version templates.
Inputs: location, project_id.
e2e_list_plansB
List available compute plans, GPU instances, CPU/RAM configurations, and pricing options.
Inputs: category, location, project_id.
e2e_list_projectsC
List all available IAM projects and access control scopes in your E2E Cloud account.
Inputs: location.
e2e_list_reserved_ipsC
List all reserved (static public) IP addresses in your project.
Inputs: location, project_id.
e2e_list_security_groupsB
List security groups attached to or available for a specific compute node.
Inputs: vm_id (required), type, location, project_id.
e2e_list_sfsC
List all SFS (Shared File System / Elastic File Storage) instances.
Inputs: location, project_id.
e2e_list_volumesA
List all block storage volumes in your E2E Cloud project (size, IOPS, attached node status, filesystem).
Inputs: page_no, location, per_page, project_id.
e2e_list_vpcsB
List all Virtual Private Clouds (VPCs) in your E2E Cloud project.
Inputs: page_no, location, per_page, project_id.
e2e_node_actionC
Execute a lifecycle action on a compute node: power_off, power_on, reboot, reinstall, save_images, rename, lock, or unlock.
Inputs: action (required), node_id (required), name, location, +1 more.
e2e_raw_requestA
Execute a direct, authenticated HTTP request against ANY E2E Networks Cloud REST API endpoint (MyAccount or TIR AI platform). Automatically handles authentication (API key + Bearer token), project scoping, and error normalization. Consult the E2E API docs (https://docs.e2enetworks.com/api/myaccount/ or https://docs.e2enetworks.com/api/tir/) for specific path structures.
Inputs: path (required), method (required), body, service, +3 more.
e2e_test_connectionA
Verify that the MCP server can authenticate and connect to the E2E Networks Cloud REST API.
Inputs: no named parameters.
e2e_tir_create_notebookC
Provision an AI Lab instance with GPU acceleration (PyTorch, TensorFlow, vLLM, etc.) in E2E TIR.
Inputs: name (required), sku_name (required), location, disk_size, +3 more.
e2e_tir_list_datasetsC
List all datasets stored in E2E TIR for AI/ML training and fine-tuning.
Inputs: location, project_id.
e2e_tir_list_gpu_skusA
List available NVIDIA GPU hardware SKUs on E2E TIR (H100, A100, L40S, A40, A30, L4, V100, T4) with pricing and specs.
Inputs: service, location, project_id.
e2e_tir_list_model_endpointsB
List deployed model inference endpoints in E2E TIR (e.g., vLLM or HuggingFace endpoints).
Inputs: location, project_id.
e2e_tir_list_notebooksC
List all AI Labs and Jupyter notebook instances in E2E TIR (AI/ML platform).
Inputs: status, page_no, location, per_page, +2 more.
e2e_tir_list_training_clustersB
List managed Slurm / distributed GPU training clusters in E2E TIR.
Inputs: location, project_id.
e2e_tir_notebook_actionC
Execute a control action (start, stop, reboot, attach_reserve_ip) on an AI Lab notebook instance.
Inputs: action (required), instance_id (required), location, project_id.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 46 tools
Most tools target a distinct resource+action (nodes, databases, k8s, volumes, buckets, VPCs, reserved IPs, TIR notebooks), so selection is generally clear. Minor overlap exists between e2e_get_node and e2e_get_node_health, and the catch-all e2e_raw_request could in principle shadow every other tool, but the descriptions keep boundaries readable.
Nearly all names follow an e2e_ prefix with a predictable verb_noun pattern (list_nodes, get_node, create_volume, delete_bucket), plus a consistent e2e_tir_ namespace for AI tools. A few names invert to noun_action (e2e_node_action, e2e_database_action, e2e_action_reserved_ip), a minor deviation but still predictable.
46 tools is heavy by count (normally warranting a low score), but the server spans a genuinely large cloud domain (compute, networking, storage, DBaaS, k8s, billing, and a full AI/ML platform). Spread across these services it is roughly 4-5 tools each, so it is borderline rather than excessive.
Several domains are read-only with no lifecycle operations: load balancers, security groups, k8s clusters, and SFS offer only list/get with no create/delete. Compute and database flows are well covered, and the e2e_raw_request escape hatch plus credential/connection tools mitigate the gaps, but the surface is uneven.
Maintenance
Related MCP Connectors
Provides capabilities that let LLM agents perform a range of infrastructure management tasks.
Your AI Agent's Infrastructure Layer. Connect Claude, Copilot, Codex, or ChatGPT to 200+ managed open source services. Start databases, pipelines, and applications through natural language.
Unified API to query AWS, GCP, Azure and generate Terraform/CLI execution kits for AI agents.
Deploy, monitor, and manage your OpenClaw AI assistants via natural language.
Related MCP Servers
- AlicenseCqualityDmaintenanceEnables AI assistants to manage multi-cloud resources (AWS, Azure, GCP) including resource operations, cost analysis, monitoring metrics, and security compliance checks through natural language commands.2613 npm2MIT
- AlicenseNot gradedqualityDmaintenanceProvides AI agents with natural language control over AWS, Azure, GCP, and Alibaba Cloud infrastructure through dynamic API discovery and execution. Supports 51,900+ cloud operations and includes OpenTofu integration for complete infrastructure lifecycle management.3MIT
- AlicenseNot gradedqualityDmaintenanceEnables AI assistants to perform DevOps tasks including Kubernetes management, cloud provider operations, CI/CD, security scanning, and infrastructure monitoring through natural language.MIT

GreenNode MCP Serverofficial
AlicenseNot gradedqualityDmaintenanceEnables AI assistants to manage VNG Cloud infrastructure including compute, storage, networking, and Kubernetes resources through natural language commands.Apache 2.0