A sophisticated audio intelligence skill for transcribing, diarizing, and extracting structured data from speech using AssemblyAI's state-of-the-art models.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install abe-audio-transcribe
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 abe-audio-transcribe 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 AssemblyAI Transcribe?
The AssemblyAI Transcribe skill is a high-performance integration for Openclaw Skills designed to handle complex speech-to-text workflows. Unlike generic transcription tools, it leverages the full power of AssemblyAI, including the universal-3-pro and universal-2 models, to provide industry-leading accuracy, automatic language detection, and speaker diarization. It is specifically engineered for AI agents, providing outputs in normalized formats like Markdown and JSON that are easy for downstream LLMs to consume.
Beyond simple transcription, this skill acts as a gateway to Speech Understanding. It supports multi-language translation, sentiment analysis, and entity extraction. By integrating the AssemblyAI LLM Gateway, users can perform structured data extraction directly from audio content, making it an essential tool for developers building automated meeting summaries, interview analyzers, or media monitoring systems within the Openclaw Skills ecosystem.
AssemblyAI Transcribe Use Cases
- Transcribing multi-speaker meetings or interviews with precise speaker labels and role mapping.
- Translating audio content into multiple languages while maintaining original utterance timing.
- Extracting structured JSON data from long-form audio using custom LLM prompts and schemas.
- Generating accessible subtitles in SRT or VTT formats for video production workflows.
- Creating agent-friendly transcript manifests for automated documentation and PKM systems.
How AssemblyAI Transcribe Works
- Input Handling: The skill accepts local file paths or public URLs; local files are securely uploaded to AssemblyAI before processing.
- Model Routing: It selects the optimal model (like universal-3-pro) based on language requirements or user preference.
- Transcription & Diarization: The audio is processed with optional features enabled, such as speaker identification and language detection.
- Post-Processing: The skill can apply speaker maps, perform translations, or run Speech Understanding tasks on the resulting transcript.
- Bundle Generation: Results are compiled into a bundle directory containing Markdown, agent-specific JSON, and raw metadata.
- LLM Integration: Optionally, the transcript is passed through the LLM Gateway for structured extraction or summarization.
AssemblyAI Transcribe Setup
Ensure you have Node.js 18+ installed and an active AssemblyAI API key. Set your API key as an environment variable to use it with Openclaw Skills.
# Set your API key
export SKILLBOSS_API_KEY='your_api_key_here'
# Run a basic transcription with all exports
node assemblyai.mjs transcribe ./audio.mp3 --bundle-dir ./output --all-exports
AssemblyAI Transcribe Data Schema & Taxonomy
The skill organizes output into a structured bundle directory for maximum scannability by AI agents and developers.
| File Type | Description |
|---|---|
transcript.md |
Human-readable Markdown with speaker labels. |
agent.json |
Normalized JSON optimized for LLM consumption. |
raw.json |
The full original response from the AssemblyAI API. |
manifest.json |
A machine-readable index of all generated assets. |
subtitles.srt/vtt |
Optional closed caption files for video players. |
AssemblyAI Transcribe Advanced Features
- LLM Gateway Integration: Use the
llmcommand to run sophisticated prompts against transcripts with strict JSON schema enforcement. - Dynamic Speaker Mapping: Apply manual speaker maps or use AssemblyAI's speaker identification to replace generic labels with real names.
- Model Routing Strategy: Automatically switch between models based on language complexity and cost requirements.
- Bundle Manifests: Generates machine-readable manifests that allow other Openclaw Skills to automatically discover and process transcript outputs.
- EU Data Residency: Support for the AssemblyAI EU base URL to meet strict regional data processing requirements.
SKILL.md
Loading
Related Openclaw Skills
RedditGrow Free: AI Agent Toolkit for Reddit
A specialized suite of six AI agents designed to automate research, content creation, and authentic community engagement on Reddit.
NLP Text Analyzer
A powerful natural language processing toolkit providing sentiment analysis, Chinese-English tokenization, and keyword extraction for AI agents.
WeShop FireRed Image Edit Skill
High-fidelity image editing and generation powered by the FireRed open-source model through the WeShop OpenAPI.
Vibevideo Core
Vibevideo Core is the migrated and updated engine for high-performance AI video generation and automation.
HK IPO Profile & Watchlist Manager
A persistent state manager for Hong Kong IPO investment profiles and symbol watchlists designed for personalized decision-making.
PDF API Hub
A comprehensive document automation integration for converting HTML to PDF, merging files, performing OCR, and managing document workflows via REST API.
Featured*
