Skip to main content
OpenClaw is an AI assistant platform that runs on your own infrastructure. When paired with StageWhisper, it receives tasks generated from your call conversations and executes them using its configured tools and capabilities.

​
Pairing process

Pairing connects your StageWhisper account with your OpenClaw host. You do this once, and the connection persists across sessions. During pairing, the two sides also exchange encryption keys for end-to-end encrypted communication.
1

Open assistant settings

In the StageWhisper desktop app, click the gear icon and go to the Assistant tab.
2

Request a pairing code

Click “Connect your AI.” StageWhisper generates a 6-digit pairing code. This code expires after about 5 minutes.
3

Enter the code in OpenClaw

Open your OpenClaw dashboard and go to the StageWhisper channel settings. Enter the pairing code.
4

Confirm the connection

StageWhisper polls for the connection. Once OpenClaw confirms the code, the pairing is complete. You’ll see a “Connected” status in the assistant settings, along with an “End-to-end encrypted” badge when BYO mode is active.

​
Connection health

After pairing, StageWhisper monitors the connection with a heartbeat:
StatusMeaning
FreshHeartbeat received within the last 90 seconds. Everything is working.
StaleNo heartbeat for 90 seconds to 5 minutes. The connection may be experiencing issues.
OfflineNo heartbeat for over 5 minutes. Tasks won’t be delivered until the connection recovers.
If the connection goes offline, StageWhisper queues approved tasks and delivers them when the connection is re-established. You don’t lose tasks.

​
What OpenClaw receives

When you approve a task during a call, StageWhisper sends a structured payload to your OpenClaw host:
  • A task description generated from the conversation context
  • The relevant transcript excerpt
  • Signal metadata (what triggered the task)
  • Playbook context (what the session is about)
Your OpenClaw host then processes the task using whatever tools and integrations you’ve configured on it (email, CRM, calendar, etc.).

​
BYO reasoning

If you have your own AI model configured on your OpenClaw host, you can switch to BYO (Bring Your Own) reasoning mode. In this mode, your call analysis runs through your own AI, on your own infrastructure.

​
How it works

Your desktop app handles speech recognition, prompt construction, and result processing locally. When it needs AI reasoning, it encrypts the analysis prompt and sends it through the StageWhisper backend to your OpenClaw plugin. Your plugin decrypts the prompt, calls your AI model, encrypts the response, and sends it back the same way. The StageWhisper backend is a blind relay. It routes encrypted messages between your desktop and your plugin but cannot read the content. See privacy and data handling for specifics on what the backend can and cannot see.

​
Trust boundaries

ComponentCan see contentRole
Your desktop appYesRuns speech recognition, builds prompts, parses results, displays signals
StageWhisper backendNoRoutes encrypted envelopes, stores session metadata
Your OpenClaw pluginYesDecrypts prompts, calls your AI, encrypts responses
BYO reasoning is available with the Founding Edition or any active subscription. See plans for details.

​
Disconnecting

You can unpair your OpenClaw host from the assistant settings at any time. This stops all task relay immediately. Your existing session data and past tasks are not affected. To re-pair, go through the pairing process again with a new code.