tescmd

Installation and setup guide for Tesla vehicle control and telemetry via the tescmd node.

Install

openclaw skills install @oceanswave/openclaw-tescmd

OpenClaw Tesla (tescmd) — Setup Guide

This plugin connects Tesla vehicles to the OpenClaw Gateway via the tescmd node. Once installed and paired, the plugin automatically registers all tools, commands, slash commands, and telemetry event types.

This document covers installation and setup only. Runtime tool usage, workflows, and error handling are provided by the tescmd skill (call tescmd_help for the full reference).

What you get:

39 agent-callable tools
14 slash commands
Real-time telemetry streaming
Supercharger discovery (10,000+ locations via supercharge.info)
CLI fallback when node is disconnected

Repositories:

Plugin: https://github.com/oceanswave/openclaw-tescmd
tescmd node (Python CLI): https://github.com/oceanswave/tescmd

Architecture

text

Agent (you)
  ↓  tool calls
OpenClaw Gateway
  ↓  node.invoke.request
openclaw-tescmd Plugin
  ↓  WebSocket dispatch
tescmd Node (Python)
  ├─ Tesla Fleet API (REST)
  ├─ Vehicle Command Protocol (VCSEC — signed commands)
  └─ Fleet Telemetry Stream (WebSocket)
  ↓
Tesla Vehicle

The plugin is the Gateway-side counterpart to the tescmd node. It defines tool schemas and routes invocations. The tescmd node handles all direct communication with Tesla.

Setup

Step 1: Check Prerequisites

Before starting, verify the required tools are installed and authenticated.

Required: git

bash

git --version

If missing, install it:

macOS: xcode-select --install
Linux: sudo apt install git or sudo dnf install git

Required: GitHub CLI (gh)

bash

gh --version
gh auth status

If gh is not installed:

macOS: brew install gh
Linux: see https://github.com/cli/cli/blob/trunk/docs/install_linux.md

If not logged in:

bash

gh auth login

Tell the user: "Please complete the GitHub CLI login in your terminal. Select your preferences when prompted and finish the browser-based auth flow."

Wait for the user to confirm they have completed the login before continuing.

Required: Python 3.11+

bash

python3 --version

Must be 3.11 or higher. If not:

macOS: brew install python@3.12
Linux: sudo apt install python3.12 or use pyenv

Recommended: Tailscale

Tailscale provides a public HTTPS endpoint for Tesla Fleet Telemetry streaming with zero infrastructure setup.

bash

tailscale version
tailscale status

If not installed:

macOS: brew install tailscale or download from https://tailscale.com/download
Linux: curl -fsSL https://tailscale.com/install.sh | sh

If not logged in:

bash

sudo tailscale up

Tell the user: "Please complete the Tailscale login in your browser if prompted."

Wait for the user to confirm before continuing.

Step 2: Install the tescmd OpenClaw Plugin

Standard install:

bash

openclaw plugins install @oceanswave/openclaw-tescmd

Verify installation:

bash

openclaw plugins list

You should see the plugin listed with version 0.9.0 (or later).

Plugin management commands:

Command	Purpose
`openclaw plugins list`	List installed plugins
`openclaw plugins info openclaw-tescmd`	Plugin details
`openclaw plugins doctor`	Check plugin health
`openclaw plugins update openclaw-tescmd`	Update to latest
`openclaw plugins enable openclaw-tescmd`	Enable the plugin
`openclaw plugins disable openclaw-tescmd`	Disable without uninstalling

Step 3: Install the tescmd CLI

bash

pip install tescmd

Verify:

bash

tescmd --version

Step 4: Run tescmd Setup

The tescmd setup wizard is interactive and requires the user to make choices and complete steps in their terminal and browser. You cannot complete this step autonomously.

bash

tescmd setup

Tell the user: "I've started the tescmd setup wizard. This is an interactive process that will walk you through:"

Creating a Tesla Developer application
Generating your EC key pair
Hosting your public key (via GitHub Pages or Tailscale Funnel)
Registering with the Tesla Fleet API
Completing OAuth2 login in your browser
Pairing the key with your vehicle (requires physical presence at the vehicle)

"Please follow the prompts in your terminal and let me know when setup is complete."

Wait for the user to confirm setup is finished before proceeding.

Verify Setup

After the user confirms, check auth status:

bash

tescmd auth status

This should show a valid token. If it shows expired or missing, the user needs to re-run:

bash

tescmd auth login

Step 5: Identify the Vehicle

List vehicles on the account to get the VIN:

bash

tescmd vehicle list

Note the VIN — it is needed for the serve command.

Step 6: Start the tescmd Node and Pair with the Gateway

The tescmd node bridges the Tesla Fleet API to the OpenClaw Gateway. The first connection requires a one-time pairing approval.

First-time pairing:

Start the node with just the Gateway URL (no token needed):

bash

tescmd serve <VIN> --openclaw <gateway_ws_url>

The node sends a node.pair.request to the Gateway and waits for approval. The pending request expires after 5 minutes, so approve it promptly.

In a separate terminal, approve the pairing:

bash

openclaw nodes pending                # View waiting pair requests
openclaw nodes approve <requestId>    # Approve the node

On approval the Gateway issues an authentication token. The node receives it, saves it to ~/.config/tescmd/bridge.json, and establishes the authenticated connection. No manual token handling is needed.

Tell the user: "Start the tescmd node with tescmd serve <VIN> --openclaw <gateway_url>, then in another terminal run openclaw nodes pending and openclaw nodes approve <requestId> to complete pairing."

Wait for the user to confirm pairing is complete before continuing.

Subsequent connections (already paired):

Once paired, the node reconnects automatically using the stored token:

bash

tescmd serve <VIN> --openclaw <gateway_ws_url>

You can also pass the token explicitly if needed:

bash

tescmd serve <VIN> --openclaw <gateway_ws_url> --openclaw-token <gateway_token>

Node management commands:

Command	Purpose
`openclaw nodes pending`	View pending pair requests
`openclaw nodes approve <id>`	Approve a node
`openclaw nodes reject <id>`	Reject a node
`openclaw nodes status`	List paired nodes and their status

Operating modes:

Mode	Command	Description
Full (default)	`tescmd serve <VIN> --openclaw <url>`	MCP server + telemetry + OpenClaw bridge
Bridge only	`tescmd serve <VIN> --no-mcp --openclaw <url>`	Telemetry + OpenClaw, no MCP server
With Tailscale	`tescmd serve <VIN> --tailscale --openclaw <url>`	Exposes MCP via Tailscale Funnel
Dry run	`tescmd serve <VIN> --dry-run`	Log events as JSONL, no Gateway connection

Key flags reference:

Flag	Description
`<VIN>`	Vehicle Identification Number (positional)
`--openclaw <ws_url>`	Gateway WebSocket URL (e.g. `ws://host:18789`)
`--openclaw-token <token>`	Gateway authentication token (auto-stored after pairing)
`--openclaw-config <path>`	Bridge config JSON (default: `~/.config/tescmd/bridge.json`)
`--transport <type>`	MCP transport: `streamable-http` (default) or `stdio`
`--port <num>`	MCP HTTP port (default: 8080)
`--host <addr>`	MCP bind address (default: 127.0.0.1)
`--telemetry-port <num>`	Telemetry WebSocket port (default: 4443)
`--fields <preset>`	Telemetry fields: `driving`, `charging`, or `all`
`--interval <sec>`	Telemetry polling interval in seconds
`--no-telemetry`	Disable telemetry streaming
`--no-mcp`	Disable MCP server
`--no-log`	Disable CSV telemetry logging
`--dry-run`	Log events as JSONL without connecting to Gateway
`--tailscale`	Expose MCP via Tailscale Funnel
`--client-id <id>`	MCP OAuth client ID
`--client-secret <secret>`	MCP OAuth client secret

Environment variables (alternative to flags):

These can be set in ~/.config/tescmd/.env:

bash

TESLA_CLIENT_ID=your-client-id
TESLA_CLIENT_SECRET=your-client-secret
TESLA_VIN=5YJ3E1EA1NF000000
TESLA_REGION=na                    # na, eu, or cn
OPENCLAW_GATEWAY_URL=ws://gateway.example.com:18789
OPENCLAW_GATEWAY_TOKEN=your-token
TESLA_COMMAND_PROTOCOL=auto        # auto, signed, or unsigned

Step 7: Verify the Connection

Once the node is running and paired, confirm it connected to the Gateway:

bash

openclaw nodes status

Or use the agent tool:

Call tescmd_node_status to check connection status

If connected, the plugin's tools are ready. Call tescmd_help for the full runtime reference including tool usage, workflows, and error handling.

Troubleshooting Setup

Problem	Solution
"no node connected"	Start the node: `tescmd serve <VIN> --openclaw <url>`
Pairing request not visible	Check `openclaw nodes pending` — requests expire after 5 minutes. Restart the node to generate a new request.
Node connects then disconnects	Check Gateway URL. Run `tescmd auth status` to verify Tesla auth.
Auth/token errors	Re-authenticate: `tescmd auth login`
Setup wizard issues	Re-run `tescmd setup` or check https://github.com/oceanswave/tescmd
Plugin not loading	Run `openclaw plugins doctor`. Check `openclaw plugins list` for the plugin entry.
Triggers say "not available"	Restart node with telemetry: remove `--no-telemetry` or add `--fields all`

Configuration

Minimal — the tescmd node handles all vehicle-specific configuration.

json

{
  "plugins": {
    "entries": {
      "openclaw-tescmd": {
        "enabled": true,
        "config": {
          "debug": false
        }
      }
    }
  }
}

CLI Quick Reference

tescmd CLI Commands

bash

tescmd serve <VIN> --openclaw <url>                              # Start node (uses stored token)
tescmd serve <VIN> --openclaw <url> --openclaw-token <token>     # Start node (explicit token)
tescmd setup                        # Interactive setup wizard
tescmd auth status                  # Check auth token status
tescmd auth login                   # Re-authenticate with Tesla
tescmd vehicle list                 # List vehicles on account
tescmd vehicle info                 # Full vehicle data snapshot
tescmd cache status                 # Check cache stats