// MCP TOOL REFERENCE

RUNPOD TOOLS

26 tools from the RunPod MCP Server, categorised by risk level.

READ TOOLS

get-container-registry-auth get-container-registry-auth get-endpoint get-endpoint get-network-volume get-network-volume get-pod get-pod get-template get-template list-container-registry-auths list-container-registry-auths list-endpoints list-endpoints list-network-volumes list-network-volumes list-pods list-pods list-templates list-templates

WRITE TOOLS

create-container-registry-auth create-container-registry-auth 3/5 create-endpoint create-endpoint 2/5 create-network-volume create-network-volume 2/5 create-pod create-pod 2/5 create-template create-template 2/5 update-endpoint update-endpoint 2/5 update-network-volume update-network-volume 2/5 update-pod update-pod 2/5 update-template update-template 2/5

DESTRUCTIVE TOOLS

delete-container-registry-auth delete-container-registry-auth 4/5 delete-endpoint delete-endpoint 4/5 delete-network-volume delete-network-volume 4/5 delete-pod delete-pod 4/5 delete-template delete-template 4/5

EXECUTE TOOLS

start-pod start-pod 3/5 stop-pod stop-pod 3/5

// FAQ

How many tools does the RunPod MCP server have? +

The RunPod MCP server exposes 26 tools across 4 categories: Read, Write, Destructive, Execute.

How do I enforce policies on RunPod tools? +

Use Intercept, the open-source MCP proxy. Write YAML rules for each tool — rate limits, argument validation, or deny rules — then run Intercept in front of the RunPod server.

What risk categories do RunPod tools fall into? +

RunPod tools are categorised as Read (10), Write (9), Destructive (5), Execute (2). Each category has a recommended default policy.

Enforce policies on RunPod

Open source. One binary. Zero dependencies.

npx -y @policylayer/intercept

github.com/policylayer/intercept →