Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
Version History
**Major update with agent-friendly outputs, model/language lookup, and advanced AssemblyAI features.** - Added scripts, reference data, sample speaker maps, and LLM extraction examples for richer AssemblyAI transcription workflows. - Introduced bundled language and model capability data (`models`, `languages` commands). - New agent-focused outputs: stable Markdown, agent JSON, and manifest for downstream automation. - Expanded support for speaker mapping, translation, subtitles, advanced post-processing, and LLM Gateway integration. - Provided detailed, workflow-oriented `SKILL.md` documentation and ready-to-use recipe examples. - All commands and exports now documented with flexible CLI usage and customisable workflows.
What This Skill Does
Transcribes audio and video using AssemblyAI, with support for speaker diarisation, named speaker mapping, language detection, and translation. Outputs include Markdown, structured JSON, subtitles, and bundle manifests designed for AI agent consumption. Post-processing options include AssemblyAI Speech Understanding and LLM Gateway for extraction and summarization.
Combines transcription, diarisation, speaker identification, translation, and LLM-based extraction in a single CLI built for scripted agent workflows rather than interactive use.
When to use it
- Transcribing recorded meetings with labeled speaker turns
- Generating subtitles for multilingual video content
- Extracting structured summaries from podcast episodes
- Identifying speakers by name or role in interview recordings
- Converting audio into agent-readable Markdown transcripts
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: Transcribe this meeting recording and label each speaker
- 1Runs the transcribe command with --speaker-labels and --bundle-dir flags on the audio file
- 2CLI uploads the file to AssemblyAI and polls until transcription completes
- 3Reads the manifest from the bundle directory to locate output files
- 4Returns the Markdown transcript with speaker-labeled turns to the user
Markdown transcript with labeled speaker turns, plus agent JSON and manifest saved to the bundle directory
Requirements
Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.