RunPod

Active

An AI agent for RunPod. It works in a scene with a live connection to RunPod. It asks before it writes, and every run is traced.

Manage RunPod GPU instances and serverless endpoints. Deploy ML models, manage GPU resources, and run inference workloads.

Infrastructure API Key 2 asset types 10 tools

What you can connect

Add these to your scene and AI gets access.

Account Accounts

A connected RunPod account

GPU Pod GPU Pods

A GPU-powered RunPod pod for AI/ML workloads

Tools

What the agent can do once this is connected.

Writes 10

Change data in the connected system.

runpod_exec Execute a shell command on the GPU pod.

runpod_write_file Write content to a file on the pod.

runpod_read_file Read a file from the pod.

runpod_list_files List files in a directory on the pod.

runpod_pod_status Get the status of the current pod.

runpod_create_pod Create a new GPU pod.

runpod_terminate_pod Terminate the current pod.

runpod_stop_pod Stop the current pod.

runpod_start_pod Start a stopped pod.

runpod_list_pods List all pods in the account.

Connect RunPod over MCP

Daslab is itself an MCP server.

Connect RunPod in a scene, then pin that scene's URL in Claude Code, Cursor, or any MCP client. The agent there gets your RunPod tools, and your approval gate and trace travel with them.

https://daslab.run/mcp/<your-world>/<your-scene>

One scene can hold RunPod and everything else you connect, so a single endpoint carries them all. Set up a client.

Use cases

Deploy and manage GPU instances
Run ML inference workloads
Monitor GPU resource utilization

From RunPod

API reference →

Similar integrations

Others in Infrastructure.

SSH Sandbox

Ready to try RunPod with Daslab?

Get started with CLI