from Terradev10
Provision and manage GPUs across clouds, create GPU Kubernetes clusters, deploy inference endpoints, and burst local compute to cloud with BYOAPI credential saf
Terradev provides natural-language driven GPU provisioning and cluster management across multiple cloud providers. It finds real‑time price quotes, provisions instances (spot/onsite), creates Kubernetes GPU clusters with sensible defaults, and deploys inference endpoints and HuggingFace Spaces.
Use Terradev when you need on‑demand GPU capacity, want to burst local workloads to cloud, create GPU K8s clusters, or deploy inference endpoints quickly while keeping API keys local (BYOAPI). Good for training, inference, and cost optimisation workflows.
Best used with agents that can execute or suggest shell/CLI commands and manage credentials locally (agents that can call the terradev CLI or produce ready-to-run bash commands).
Terradev is a cross-cloud GPU provisioning skill that wraps a CLI for quoting, provisioning, and managing GPU instances across 11+ cloud providers. The SKILL.md is well-structured with clear CLI commands, realistic examples, and safety-first defaults (dry-run, BYOAPI). No bundled scripts were present, so only static analysis was performed. The skill follows the spec well with complete frontmatter and progressive disclosure from simple quotes to complex K8s cluster creation.
Clean skill, no security concerns. BYOAPI approach is well-designed. Could improve with an error-handling section (what to do when provision fails, when instances are unreachable, etc.). The CLI tool itself (terradev-cli) was not tested — only the SKILL.md was reviewed. Pricing context section is a nice touch for user awareness.
