Openclaw Skills for Pod AI lets you securely inspect schemas and run Pod AI actions through the oo CLI without handling raw tokens.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install oo-pod-ai
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 oo-pod-ai 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 Pod AI Connector?
Pod AI Connector is an Openclaw Skills workflow for operating Pod AI through an OOMOL-connected account using the oo CLI. It is designed to help agents safely read, create, and update Pod AI data by fetching live action schemas first and then executing validated payloads against the pod_ai connector.
The key value of this Openclaw Skills package is that OOMOL injects credentials server-side, so you never need to manage API keys directly. This makes Pod AI automation more reliable, auditable, and agent-friendly, while still preserving safety controls for write and destructive actions.
Pod AI Connector Use Cases
- Trigger Pod AI operations from an AI coding agent without calling the API directly.
- Inspect the live contract for a Pod AI action before constructing a request payload.
- Create an outbound phone call with a Pod AI voice agent.
- Run read or list actions safely when no write or destructive state change is involved.
- Recover from setup issues like missing CLI installation, expired credentials, or disconnected app connections.
- Handle billing or credit interruptions before retrying failed Pod AI runs.
How Pod AI Connector Works
- Verify the target Pod AI action and determine whether it is a read, write, or destructive operation.
- Inspect the authoritative input/output schema with
oo connector schemafor the selected action. - Build a JSON payload that matches the live schema exactly.
- Run the action through the
pod_aiconnector withoo connector run, passing the payload via--data. - Read the result from the returned JSON, including
dataandmeta.executionIdfor traceability. - If a command fails due to setup issues, perform the matching first-time setup step only once, such as installing
oo, signing in, reconnecting Pod AI, or resolving billing credit errors.
Pod AI Connector Setup
- Install the
ooCLI if it is missing.
curl -fsSL https://cli.oomol.com/install.sh | bash
For Windows PowerShell:
irm https://cli.oomol.com/install.ps1 | iex
- Sign in to your OOMOL account if authentication is required.
oo auth login
- Connect Pod AI in the OOMOL console if the connector is missing, expired, or lacks scope.
https://console.oomol.com/app-connections?provider=pod_ai
- Confirm billing credit if requests fail with HTTP 402 or
OOMOL_INSUFFICIENT_CREDIT.
https://console.oomol.com/billing/token-recharge
- Before running any action, inspect the live schema.
oo connector schema "pod_ai" --action "<action_name>"
- Execute the action with a validated JSON payload.
oo connector run "pod_ai" --action "<action_name>" --data '{"key":"value"}' --json
Pod AI Connector Data Schema & Taxonomy
Pod AI is organized around connector actions, each with its own live input/output contract that must be inspected before execution. Openclaw Skills treats the schema as the source of truth for payload construction.
Core execution data
| Item | Purpose | Source |
|---|---|---|
action_name |
Identifies the Pod AI operation to run | User intent / connector action list |
--data |
JSON payload sent to the action | CLI input or @path/to/file.json |
data |
Returned response payload | oo connector run output |
meta.executionId |
Execution trace identifier | Response metadata |
Action taxonomy
| Tag | Meaning | Behavior |
|---|---|---|
read |
Untagged actions | Safe to run directly |
[write] |
Mutates Pod AI state | Confirm exact payload and effect first |
[destructive] |
Overwrites or removes data | Require explicit approval before running |
File and payload conventions
- Payloads may be passed inline as a JSON string.
- Payloads may also be loaded from a JSON file using
--data @path/to/file.json. - If
--datais omitted, an empty object{}is sent. - Always fetch the current schema before building a request, because the live connector contract is authoritative.
Supported action inventory
create_outbound_call— create an outbound phone call with a Pod AI voice agent.[write]
Pod AI Connector Advanced Features
- Live schema introspection with
oo connector schemafor contract-first execution. - Secure server-side credential injection through OOMOL, eliminating direct token handling.
- Support for JSON payloads inline or from files, making Openclaw Skills easy to automate in scripts and agents.
- Execution metadata via
meta.executionIdfor traceability and debugging. - Safety-aware tagging for read, write, and destructive actions to reduce accidental state changes.
- Recovery paths for CLI installation issues, authentication failures, connector expiration, scope problems, and billing interruptions.
SKILL.md
Loading
Related Openclaw Skills
Minerstat
Openclaw Skills for Minerstat lets you safely read coin, hardware, and pool data through the oo CLI without handling raw credentials.
Octopus Deploy
Openclaw Skills for Octopus Deploy lets you safely inspect deployments, releases, environments, projects, spaces, and tasks through the oo CLI.
Postalytics
Openclaw Skills for Postalytics provides schema-driven CLI access to read campaigns, contacts, templates, and stats through a connected account.
Cloudflare Docs
Openclaw Skills for Cloudflare Docs lets AI agents search, read, and safely execute documentation-backed actions through the oo CLI.
LIFX Control via OOMOL
Openclaw Skills for LIFX lets you manage lights, scenes, power, colors, and effects through the OOMOL-connected oo CLI without handling raw API tokens.
Leonardo.Ai Connector Skill
Openclaw Skills for Leonardo.Ai provide a schema-driven CLI workflow to inspect contracts, run actions, and manage generations through a connected OOMOL account.
Featured*
