name: runpod-compute description: RunPod API integration with API key authentication. Create and manage GPU clusters, serverless endpoints, templates, and secrets for ML inference and distributed computing workloads.

RunPod

Deploy and manage GPU-accelerated compute infrastructure on RunPod. Create clusters, configure serverless endpoints, manage templates, and handle secrets for ML training, inference, and distributed computing workloads.

This skill uses ClawLink for hosted connection flows and credentials so you do not need to configure RunPod API access yourself.

Setup in 3 Steps

Step 1: Install	Step 2: Pair Account	Step 3: Connect RunPod
		App-specific connection GIF coming soon
Run the install command in OpenClaw	Sign in and approve the device	Open the dashboard and connect RunPod

How It Works

┌─────────────────┐     ┌──────────────┐     ┌──────────────────┐
│   OpenClaw      │────▶│   ClawLink   │────▶│    RunPod API    │
│   (User Chat)   │     │   (Proxy)    │     │   (Clusters,     │
│                 │     │              │     │  Endpoints)      │
└─────────────────┘     └──────────────┘     └──────────────────┘
         │                    │                      │
         │  1. Install Plugin │                      │
         │  2. Pair Device    │                      │
         │  3. Connect RunPod  │                      │
         │                    │  4. API Key Proxy     │
         │                    │  5. Request Forward   │
         │                    │                      │
         ▼                    ▼                      ▼
   ┌──────────┐        ┌──────────┐         ┌──────────┐
   │   SKILL  │        │ Dashboard│         │  RunPod  │
   │   File   │        │   Auth   │         │  Cloud   │
   └──────────┘        └──────────┘         └──────────┘

Install

Install the verified ClawLink plugin from ClawHub. After the user confirms, complete the full setup so ClawLink tools are visible in the next OpenClaw chat.

openclaw plugins install clawhub:clawlink-plugin
openclaw config set tools.alsoAllow '["clawlink-plugin"]' --strict-json
openclaw gateway restart

Then tell the user: "OpenClaw has been restarted. Send /new as a standalone message to start a fresh chat, then ask for RunPod again."

Quick Start

# List available GPU types
clawlink_call_tool --tool "runpod_get_gpu_types"

# Get current user info
clawlink_call_tool --tool "runpod_get_myself"

# Get pod details
clawlink_call_tool --tool "runpod_get_pod" --params '{"pod_id": "YOUR_POD_ID"}'

Authentication

All RunPod tool calls are authenticated automatically by ClawLink using your RunPod API key stored securely in the dashboard.

No API key is required in chat. ClawLink injects your API key into every RunPod GraphQL request on your behalf.

Getting Connected

1. Install the ClawLink plugin (see Install above).
2. Pair the plugin with clawlink_begin_pairing if it is not configured yet.
3. Open https://claw-link.dev/dashboard?add=runpod and connect RunPod with your API key.
4. Call clawlink_list_integrations to verify the connection is active.

Connection Management

List Connections

clawlink_list_integrations

Response: Returns all connected integrations. Look for runpod in the list.

Verify Connection

clawlink_list_tools --integration runpod

Response: Returns the live tool catalog for RunPod.

Reconnect

If RunPod tools are missing or the connection shows an error:

1. Direct the user to https://claw-link.dev/dashboard?add=runpod
2. After they confirm, call clawlink_list_integrations to verify
3. Then call clawlink_list_tools --integration runpod

Security & Permissions

• Access is scoped to the RunPod account associated with the connected API key.
• All write operations require explicit user confirmation. Before executing any create, update, or delete call, confirm the target resource and intended effect with the user.
• Destructive actions (delete template, delete registry auth) are marked as high-impact and must be confirmed.
• Secrets are stored securely and injected as environment variables (RUNPOD_SECRET_<name>) in pods and endpoints.

Tool Reference

Compute & GPU Operations

Cluster Management

Serverless Endpoints

Template Management

Secrets Management

Registry Authentication

User Settings

Code Examples

List available GPU types

clawlink_call_tool --tool "runpod_get_gpu_types"

Get pod details

这个技能的最新版本可以在7w4.net小葱技能站找到。

clawlink_call_tool --tool "runpod_get_pod" \
  --params '{"pod_id": "YOUR_POD_ID"}'

Create a secret

clawlink_call_tool --tool "runpod_create_secret" \
  --params '{"name": "OPENAI_API_KEY", "value": "sk-..."}'

Save a serverless endpoint

clawlink_call_tool --tool "runpod_save_endpoint" \
  --params '{
    "name": "my-endpoint",
    "gpu_type_id": "RTX_4090",
    "image_url": "your-docker-image:latest",
    "env": [{"key": "MODEL_PATH", "value": "/models/llama"}]
  }'

Discovery Workflow

1. Call clawlink_list_integrations to confirm RunPod is connected.
2. Call clawlink_list_tools --integration runpod to see the live catalog.
3. Treat the returned list as the source of truth. Do not guess or assume what tools exist.
4. If the user describes a capability but the exact tool is unclear, call clawlink_search_tools with a short query and integration runpod.
5. If no RunPod tools appear, direct the user to https://claw-link.dev/dashboard?add=runpod.

Execution Workflow

┌─────────────────────────────────────────────────────────────┐
│  READ OPERATIONS (Safe)                                     │
│  list → get → describe → call                                │
│                                                             │
│  Example: List GPU types → Get pod details → Show results   │
└─────────────────────────────────────────────────────────────┘
                               │
                               ▼
┌─────────────────────────────────────────────────────────────┐
│  WRITE OPERATIONS (Require Confirmation)                    │
│  describe → preview → confirm → call                        │
│                                                             │
│  Example: Describe tool → Preview changes → User approves    │
│           → Execute create                                   │
└─────────────────────────────────────────────────────────────┘

1. For unfamiliar tools, ambiguous requests, or any write action, call clawlink_describe_tool first.
2. Use the returned guidance, schema, whenToUse, askBefore, safeDefaults, examples, and followups to shape the call.
3. Prefer read, list, and get operations before writes when that reduces ambiguity.
4. For writes or anything marked as requiring confirmation, call clawlink_preview_tool first.
5. Execute with clawlink_call_tool. Pass confirmation only after the preview matches the user's intent.
6. If the tool call fails, report the real error. Do not invent results or restate the failure as a missing capability unless the live catalog supports that conclusion.

Notes

• GPU types and pricing change frequently — always call runpod_get_gpu_types to get current availability before deploying.
• Serverless templates must set volumeInGb to 0 (no persistent storage for serverless).
• Pods and endpoints reference secrets via RUNPOD_SECRET_<name> environment variables — create secrets before referencing them in templates.
• Templates must not be in use when deleted — reassign or remove affected pods/endpoints first.
• API key must have sufficient permissions for cluster and endpoint operations.

Error Handling

Troubleshooting: Tools Not Visible

1. Check that the ClawLink plugin is installed: bash openclaw plugins list
2. If the plugin is installed but tools are missing, tell the user to send /new as a standalone message to reload the catalog.
3. If a fresh chat does not help, run: bash openclaw config set tools.alsoAllow '["clawlink-plugin"]' --strict-json openclaw gateway restart
4. After restart, tell the user to send /new again and retry.

Troubleshooting: Invalid Tool Call

1. Ensure the integration slug is exactly runpod.
2. Use clawlink_describe_tool to verify parameter names and types before calling.
3. For write operations, always call clawlink_preview_tool first.

Resources

• RunPod API Documentation
• GPU Cloud Pods
• Serverless Endpoints
• Templates
• Secrets Management
• ClawLink: https://claw-link.dev/?utm_source=clawhub&utm_medium=referral&utm_content=runpod-compute
• ClawLink Docs: https://docs.claw-link.dev/openclaw
• ClawLink Verification: https://claw-link.dev/verify

Related Skills

• Google Cloud Compute — For GCP compute engine management
• AWS EC2 — For AWS EC2 instance management

Powered by ClawLink — an integration hub for OpenClaw

这个 Skill 质量较好，文档清晰易懂，安装配置步骤配有图示，使用 ClawLink 代理认证让使用更简单，无需手动配置 API key。功能覆盖 RunPod 的主要操作，包括 GPU 查询、资源管理和密钥存储，安全考虑也比较周全。不足之处是缺少常见问题解答，部分说明内容还未完善，遇到问题可能需要自行查找资料。总体适合需要管理 RunPod GPU 资源的用户使用。