A cross-cloud GPU orchestration agent that enables real-time pricing comparison, automated provisioning, and NUMA-optimized Kubernetes cluster creation across 11+ providers.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install terradev-gpu-cloud
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install terradev-gpu-cloud using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Terradev GPU Cloud?
Terradev GPU Cloud is a comprehensive infrastructure-as-code solution designed specifically for AI developers and DevOps engineers. By leveraging Openclaw Skills, users can bypass the complexity of individual cloud consoles to find the cheapest H100, A100, or consumer-grade GPUs in seconds. The skill operates on a Bring Your Own API (BYOAPI) model, ensuring that sensitive credentials never leave the user's local environment while providing a unified interface for providers like RunPod, Lambda Labs, AWS, GCP, and Azure.
Beyond simple instance launching, this skill automates the deployment of production-grade Kubernetes clusters and serverless inference endpoints. It handles the intricate technical details of high-performance computing, such as GPU-NIC pairing and NUMA-aligned topology optimization, ensuring that your AI workloads run with maximum inter-node bandwidth and minimum latency. This integration with Openclaw Skills represents a significant shift toward agentic infrastructure management.
Terradev GPU Cloud Use Cases
- Comparing real-time spot and on-demand GPU prices across 11+ cloud providers simultaneously.
- Provisioning high-performance GPU instances for LLM training with a strict budget ceiling.
- Deploying multi-cloud Kubernetes clusters with Karpenter-driven spot-first scheduling.
- Offloading local machine workloads to cloud GPUs when VRAM limits are exceeded.
- Serving HuggingFace models via serverless InferX endpoints for production-ready inference.
How Terradev GPU Cloud Works
- The user requests a GPU price quote or provisioning plan via Openclaw Skills using natural language.
- The underlying CLI queries multiple cloud provider APIs in parallel to fetch real-time availability and technical specs.
- A detailed plan is generated, presenting the user with the most cost-effective hardware options and locations.
- Upon approval, the skill executes the provisioning lifecycle, including network configuration and hardware-level optimizations.
- For Kubernetes deployments, the agent configures NUMA alignment and CNI-first ordering to avoid common race conditions.
- The user manages the infrastructure lifecycle—from execution to termination—through a centralized command set.
Terradev GPU Cloud Setup
To begin using this skill with Openclaw Skills, install the Terradev CLI and configure your local environment variables:
# Install the CLI with all provider SDKs
pip install "terradev-cli[all]"
# Configure your primary provider (e.g., RunPod)
export TERRADEV_RUNPOD_KEY=your_api_key_here
# (Optional) Configure AWS for multi-cloud capabilities
export TERRADEV_AWS_ACCESS_KEY_ID=your_access_key
export TERRADEV_AWS_SECRET_ACCESS_KEY=your_secret_key
You can verify your setup by running terradev status --live to check connectivity to your configured providers.
Terradev GPU Cloud Data Schema & Taxonomy
The skill manages infrastructure metadata locally to ensure privacy and speed. Below is how data is structured:
| Component | Data Managed | Description |
|---|---|---|
| Quotes | Pricing, Region, GPU Type | Real-time market data from 11+ providers. |
| Instances | ID, IP, SSH Keys, State | Tracking for active compute resources. |
| Clusters | Kubeconfig, NodeClass, CNI | Configuration for NUMA-optimized K8s environments. |
| Analytics | Spend, Hours, Provider Mix | Cost management and optimization logs. |
Terradev GPU Cloud Advanced Features
- NUMA-Aligned Topology: Automatic placement of GPUs and network cards behind the same PCIe switch to eliminate cross-socket latency.
- InferX Serverless: Deploy models directly to a serverless platform with automated hardware scaling and cost analysis.
- Karpenter Integration: Advanced Kubernetes scheduling that prioritizes spot instances while maintaining cluster stability.
- Local-to-Cloud Bursting: Seamlessly move Dockerized workloads from local hardware to cloud instances with the
runcommand. - Multi-Cloud Node Pools: Orchestrate a single Kubernetes cluster spanning AWS, GCP, and CoreWeave for maximum reliability.
SKILL.md
Loading
Related Openclaw Skills
mdnew
A high-efficiency conversion tool that transforms web URLs into clean, token-optimized Markdown for AI agents.
Kradleverse Observe
A specialized skill for capturing and analyzing real-time observations from within a Kradleverse environment.
Kradleverse AI Minecraft Agent
An autonomous AI agent skill for playing multiplayer Minecraft via the Kradleverse platform.
Kradleverse Observe
A powerful utility for extracting real-time environmental observations and game state data for AI agents.
PostCraft
PostCraft is an automated content polisher that transforms raw text into structured, high-engagement MoltBook posts using professional markdown standards.
PostCraft
PostCraft is an automated formatting engine designed to polish MoltBook posts into professional, highly readable Markdown documents.
Featured*
