A voice-interface skill that enables ambient wake word detection and command execution for OpenClaw agents.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install percept-voice-cmd
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 percept-voice-cmd 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 Percept Voice Command?
Percept Voice Command is a specialized tool designed to bring ambient voice intelligence to your workspace. By acting as a bridge between spoken language and agent execution, it allows users to interact with their digital environment without manual input. This skill is a core component for those looking to expand their library of Openclaw Skills with voice-activated capabilities.
The system monitors ambient speech for specific wake words and instantly routes captured commands to the underlying agent. Whether you are managing communications or querying data, this skill ensures your AI is always listening and ready to assist through the robust Openclaw Skills framework.
Percept Voice Command Use Cases
- Hands-free agent control during focused work sessions.
- Quick voice-to-text messaging and email drafting.
- Setting instant reminders and calendar events while away from the keyboard.
- Managing notes and server information via custom voice triggers.
How Percept Voice Command Works
- The skill monitors incoming audio transcript segments provided by the percept-listen dependency.
- When a predefined wake word is detected, the capture buffer is extended by 5 seconds to ensure the full command is recorded.
- A 10-second continuation window is opened after the initial command to allow for natural follow-up interactions without repeating the wake word.
- The system employs a two-tier parsing logic, starting with fast regex for common tasks and falling back to an LLM for complex intent resolution.
- Contact details are resolved using local data files before the final command is dispatched to the agent CLI for execution.
Percept Voice Command Setup
To get started with this addition to your Openclaw Skills collection, ensure the requirements are met and run the following commands:
# Ensure percept-listen is installed and active
# Install the voice command skill
openclaw install percept-voice-cmd
Access the Percept dashboard at http://localhost:8960 to configure your specific wake words and authorized speakers.
Percept Voice Command Data Schema & Taxonomy
The skill organizes its metadata and authorization protocols through local JSON structures and a centralized database to ensure fast response times for all Openclaw Skills.
| Data Component | Format | Purpose |
|---|---|---|
contacts.json |
JSON | Maps names to email addresses and phone numbers for command resolution. |
speakers.json |
JSON | Tracks speaker IDs and authorization levels to secure command execution. |
| Wake Word DB | SQL/Key-Value | Stores primary triggers like 'Hey Jarvis' or custom action phrases. |
Percept Voice Command Advanced Features
- Intelligent speaker authorization that restricts command execution to approved users only.
- Hybrid intent parsing utilizing both regex and LLM fallback for maximum efficiency.
- Seamless contact resolution from local address books for personalized messaging.
- Support for custom wake word configuration via a dedicated web dashboard.
- Extensible command routing that supports the entire ecosystem of Openclaw Skills.
SKILL.md
Loading
Related Openclaw Skills
percept-summarize
An automated conversation intelligence tool that generates AI summaries and maps entity relationships from transcribed voice data.
Percept Speaker ID
A sophisticated speaker identification system that maps anonymous transcript labels to real names and manages voice command authorization levels.
Percept Listen
Ambient audio capture and local transcription for AI agents via wearable devices like Omi and Apple Watch.
percept-ambient
An always-on ambient intelligence layer that builds a continuous knowledge graph of your conversations and projects without needing explicit commands.
Stable Browser CDP
A high-stability browser automation layer using Chrome DevTools Protocol to replace unreliable extension-based relays.
Hypii Hyperliquid Trading Agent
An AI-powered trading agent for Hyperliquid perpetual futures featuring automated strategies and integrated micropayments.
Featured*
