spaces-listener

What This Skill Does

The spaces-listener skill is a powerful automation tool for OpenClaw agents designed to capture, process, and distill information from X (Twitter) Spaces. Whether you are tracking live events, interviews, or casual discussions, this skill handles the heavy lifting of audio extraction, high-fidelity transcription via OpenAI's Whisper, and intelligent summarization. It manages the entire lifecycle of a Space, from initiating a background recording session that persists independently of your terminal to organizing the resulting audio, logs, transcripts, and summaries into a structured local directory.

Installation

To get started, ensure you have the necessary system dependencies installed via Homebrew: brew install yt-dlp ffmpeg openai-whisper

Next, install the skill directly into your OpenClaw environment: clawhub install openclaw/skills/skills/jamesalmeida/spaces-listener

For summarization capabilities, configure your environment by setting your OPENAI_API_KEY. Without this key, the tool will still perform transcription but will skip the generation of AI-driven summary reports.

Use Cases

• Journalism and Research: Automatically archive expert commentary and live discussions from public figures or industry panels for later analysis.
• Content Creation: Turn long-form audio discussions into summarized blog posts, threads, or newsletter content by utilizing the automated transcript generation.
• Meeting Intelligence: Record and summarize internal or team-hosted Spaces to ensure key takeaways are documented without requiring manual note-taking.
• Accessibility: Provide text-based transcripts for audio-only content, making information shared in Spaces searchable and accessible for those who prefer reading.

Example Prompts

• "OpenClaw, listen to this Space URL and notify me once it has been transcribed and summarized: [Insert URL]"
• "List all my currently active space recordings and give me a status update on recording #3."
• "Stop all current recordings and delete the stale metadata files to clean up my storage."

Tips & Limitations

This skill is designed for long-duration audio; therefore, never impose a timeout on the download process, as the listener is architected to finish only when the Space concludes naturally. Use the provided cron job strategy to monitor the status of background recordings effectively. Note that transcription accuracy depends on the chosen Whisper model; use 'medium' or 'large' for complex discussions, or 'tiny'/'base' for quick, high-speed turnarounds. All files are staged in /tmp/ before being moved to your Dropbox/ClawdBox/XSpaces/ folder to prevent Dropbox sync conflicts during the writing process.