A high-precision technical skill for generating professional AI voiceovers with automated scene analysis and frame-accurate audio-video synchronization.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install narrate-video
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 narrate-video 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 Video Narration?
The Video Narration skill is a sophisticated tool within the Openclaw Skills ecosystem designed for developers and content creators who need to add high-quality commentary to video files. By leveraging Azure Text-to-Speech (TTS) and FFmpeg, it automates the tedious process of timing scripts to visual scene changes. It doesn't just generate audio; it analyzes the underlying video stream to ensure that narration begins only after relevant visual content is visible, creating a natural presenter-like experience.
This skill is particularly optimized for technical demos, screen recordings, and educational tutorials. It provides a robust framework for managing speech rates, calculating word counts per time window, and handling complex audio mixing challenges like volume normalization and multi-segment overlapping. Through a modular Python-based workflow, Openclaw Skills enables users to iterate on scripts and timing until the final output is polished and professionally synced.
Video Narration Use Cases
- Creating narrated tutorials or walk-throughs for software products
- Adding professional commentary to recorded presentations or webinars
- Localizing video content with dubbed voiceovers in multiple languages
- Generating automated explanations for complex screen recordings or bug reports
- Transforming silent video demos into engaging marketing material
How Video Narration Works
- The skill performs a deep analysis of the video metadata and extracts frames at regular intervals to map scene transitions.
- Users define or refine a script where narration segments are mapped to specific start-time timestamps.
- The system calculates word counts based on speech rates to ensure the text fits within the available visual window without rushing.
- Individual audio segments are generated via Azure TTS and cached to allow for efficient iteration.
- FFmpeg is used to mix the generated audio segments into a single track, carefully synchronized with the original video stream.
- A final verification phase checks for volume levels, silence gaps, and visual alignment to guarantee a professional finish.
Video Narration Setup
Ensure you have FFmpeg and Python installed on your system. You will also need an active Azure Speech service subscription.
# Install the Azure Cognitive Services Speech SDK
pip install azure-cognitiveservices-speech python-dotenv
# Configure your environment credentials
echo "AZURE_SPEECH_KEY=your-key-here" >> ~/.narrate_video.env
echo "AZURE_SPEECH_REGION=your-region-here" >> ~/.narrate_video.env
# Verify dependencies are accessible
command -v ffmpeg && command -v ffprobe
Video Narration Data Schema & Taxonomy
The skill maintains a structured workflow using a dedicated Python script and a directory for cached audio segments.
| File/Folder | Description |
|---|---|
narration_script.py |
The main execution script containing segments, voice settings, and file paths. |
narration_segments/ |
Directory containing individual .mp3 files for each narration segment. |
output_video.mp4 |
The final rendered video with the merged narration track. |
references/voices.md |
A lookup table for supported Azure voices and their estimated speech rates. |
Video Narration Advanced Features
- Smart Caching: Only regenerates audio for segments where the text has changed, saving API costs and time.
- Overlap Detection: Automatically aborts the build process if audio segments overlap, preventing broken transitions.
- Volume Normalization: Uses specialized FFmpeg filters to ensure consistent audio levels across all mixed segments.
- Variable Speech Rates: Supports fine-tuning of word-per-second estimates based on the specific Azure voice selected.
- Visual-First Logic: Enforces a delay between visual appearance and audio start to mimic natural human presentation styles.
SKILL.md
Loading
Related Openclaw Skills
Iberia Insight Agent
A specialized research skill for deep-diving into Iberia brand history, product ecosystems, and global market positioning.
DataWorks Diagnoser
A specialized CLI tool for fetching and diagnosing Alibaba Cloud DataWorks task instance logs to identify and resolve execution failures.
Memory Focal System
An intelligent memory management framework that optimizes AI context by dynamically adjusting memory focal depth and applying tiered storage.
iHeartRadio Intelligence Skill
A comprehensive knowledge skill providing deep-dive insights into the iHeartRadio brand ecosystem and the broader streaming media industry.
nova-memory
A local, pure-Python semantic memory engine that provides AI agents with persistent entity tracking, relationship mapping, and automated reflection.
Humana Intelligence & Research Skill
An AI agent skill designed to provide structured intelligence, business maps, and market performance data for the Humana organization.
Featured*
