A powerful macOS command-line interface for transcribing audio and video files using local Whisper models or high-speed cloud processing.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install whisnap
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 whisnap 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 whisnap?
Whisnap is a specialized macOS CLI utility designed to bring robust speech-to-text capabilities to the terminal. It leverages the Whisper architecture to provide highly accurate transcriptions of both audio and video files. By integrating whisnap into your library of Openclaw Skills, you can automate transcription tasks without leaving your development environment, choosing between the privacy of local models or the speed of cloud-based processing.
The tool is built to work in tandem with the Whisnap desktop application, reusing its models and settings to ensure a consistent experience. It is particularly valuable for developers building automated media pipelines, researchers processing large volumes of recorded data, and power users who prefer the speed and flexibility of a command-line interface over a GUI.
whisnap Use Cases
- Transcribing meeting recordings or lectures directly from the terminal for quick documentation.
- Automating the generation of subtitles and timestamps for video content in media workflows.
- Batch processing audio archives using local Whisper models to maintain data privacy.
- Integrating speech-to-text functionality into custom scripts and Openclaw Skills.
- Extracting structured JSON data from media files for use in searchable databases.
How whisnap Works
- The user initiates a transcription by passing a media file path to the whisnap CLI.
- The CLI checks for the presence of the Whisnap macOS app and its configuration settings.
- Depending on the flags used, the tool selects either a locally downloaded Whisper model or connects to Whisnap Cloud.
- The audio or video stream is processed, converting speech into text with associated metadata.
- The transcription is returned as standard output or formatted as a structured JSON object for programmatic handling.
whisnap Setup
To integrate this tool with your Openclaw Skills, follow these installation steps:
- Install the Whisnap macOS application.
- Navigate to Settings > Advanced and select Enable CLI to create the
/usr/local/bin/whisnapsymlink. - Download at least one Whisper model within the app (e.g., small or medium).
- If using cloud features, ensure you are signed in through the desktop app interface.
Verify your setup by listing available models:
whisnap --list-models
whisnap Data Schema & Taxonomy
When the --json flag is utilized, whisnap generates structured metadata that is highly compatible with other Openclaw Skills. The schema includes:
| Property | Type | Description |
|---|---|---|
text |
String | The complete transcribed text of the media file. |
segments |
Array | Objects containing start_ms, end_ms, and the specific text for that segment. |
model |
String | The ID of the Whisper model used for the transcription. |
backend |
String | The processing engine used (e.g., whisper or cloud). |
processing_time_ms |
Number | The total time taken to process the file in milliseconds. |
whisnap Advanced Features
- Local-First Privacy: Perform full transcriptions locally without an internet connection using downloaded Whisper models.
- Cloud Acceleration: Use the
--cloudflag to offload heavy processing tasks to Whisnap's high-performance servers. - Model Customization: Target specific models using the
--modelflag to balance speed and accuracy. - Developer-Friendly Output: Pipe structured JSON output directly into other CLI tools or data processing scripts.
- Verbose Monitoring: Access real-time diagnostics and progress reporting using the
--verboseflag for complex debugging.
SKILL.md
Loading
Related Openclaw Skills
SONA
SONA is an autonomous Solana wallet agent that combines AI-driven reasoning with Rust-enforced safety protocols to manage on-chain transactions securely.
OpenClaw Session Cleanup
A maintenance and stability skill designed to prune stale sessions and optimize resource usage for long-running OpenClaw deployments.
Doppel
Transform WhatsApp chat history into high-fidelity AI personas that mimic real human communication styles and personalities.
DuckDuckGo Web Search
A privacy-focused web search skill that allows AI agents to query DuckDuckGo and retrieve real-time information without requiring an API key.
ClawdWallet
A multi-chain Web3 wallet Chrome extension that allows AI agents to manage crypto assets and sign transactions across 20+ blockchains.
Ghost Browser
An automated Chrome browser controlled via CLI, optimized for AI agents to perform web tasks with minimal token consumption.
Featured*
