skills.rest
framerslab
framerslabOFFICIAL
Explore all skills by this author →
Website𝕏@framerslabGitHub

porcupine

Detect configured wake-words on-device from 16 kHz mono LINEAR16 PCM frames.

1|Updated Feb 8, 2026
Search Tags:#privacy#wake-word#hotword#porcupine#voice-activation#picovoice
One-click install
npx skills add https://github.com/framerslab/agentos-skills --skill porcupine
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: porcupine
Source: https://github.com/framerslab/agentos-skills/tree/main/registry/curated/porcupine
Command: npx skills add https://github.com/framerslab/agentos-skills --skill porcupine

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires picovoice, and includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill provides on-device wake-word detection, addressing the need for privacy-preserving and configurable hands-free voice agent activation.

Core Features & Use Cases

  • On-device Wake-word Detection: Detects wake-words like "porcupine" without sending audio to external servers.
  • Configurable Keywords: Allows setting custom wake-words and sensitivity levels.
  • Use Case: Use this Skill to start a voice agent only when a specific wake-word is spoken, enhancing privacy and reducing power consumption.

Quick Start

Set up the porcupine skill with the wake-word "porcupine" and configure it to listen for the keyword.

Frequently Asked Questions about porcupine

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I enable on-device wake-word detection for a voice agent?▼

On-device wake-word detection is enabled by processing 16 kHz mono LINEAR16 PCM frames through the Picovoice Porcupine engine to trigger voice agent activation locally without external servers.

What do I need to configure custom wake-words for privacy-conscious hands-free activation?▼

Configuring custom wake-words requires a Picovoice access key and an input stream of 16 kHz mono LINEAR16 PCM audio frames to detect specified keywords with adjustable sensitivity levels.

Does Picovoice wake-word detection work without sending audio data to external servers?▼

Yes, Picovoice wake-word detection operates entirely on-device, processing audio frames locally to ensure privacy-preserving voice activation without transmitting audio data to external servers.

How can I tune the sensitivity of hotword detection to reduce false triggers?▼

Hotword detection sensitivity can be tuned by adjusting the sensitivity levels configured within the Porcupine engine to balance accurate wake-word recognition and reduce false triggers.

What audio format is required for Porcupine wake-word detection?▼

Porcupine wake-word detection requires 16 kHz mono LINEAR16 PCM audio frames to accurately process the input stream and detect configured keywords for voice agent activation.

Related Skills

synqingsynqing
15

SPH0645-Microphone-Integration

Convert SPH0645 PDM audio to PCM data for ESP32 DSP processing.

Community
Intermediate
vignesh2027vignesh2027
6

voice-agent-builder

Orchestrate streaming voice AI workflows with ASR, LLM, and TTS integration.

Community
Advanced
framerslabframerslab

voice-conversation

Enable live voice conversations with speech-to-text, text-to-speech, and wake-word activation.

Official
Advanced
ai-enhanced-engineerai-enhanced-engineer
3

gcp-cicd-patterns

Automate CI/CD pipeline design on Google Cloud Platform with Cloud Build and Cloud Deploy.

Official
Advanced