loom

What This Skill Does

The Loom skill for OpenClaw empowers your AI agent to interact seamlessly with your Loom video library. By integrating the Loom REST API, this skill allows users to list recordings, retrieve detailed metadata, fetch auto-generated transcripts, update video settings, and delete outdated content. Furthermore, it provides access to video insights, allowing you to track engagement metrics directly through your agent. It serves as a command-line interface proxy to manage your asynchronous video communications.

Installation

To integrate this skill, use the following command in your terminal within the OpenClaw environment:

clawhub install openclaw/skills/skills/mrgoodb/loom

After installation, you must configure your API credentials. Obtain your API key from the Loom developer portal and set it as an environment variable:

export LOOM_API_KEY="your_api_key_here"

Use Cases

This skill is ideal for teams heavily reliant on async video for collaboration. Use it to quickly locate a specific project update, summarize long video transcripts into concise text notes, clean up old recordings to manage storage, or pull engagement analytics to verify if your team has viewed critical training documentation.

Example Prompts

1. "List my last 5 Loom videos and tell me which ones have more than 10 views."
2. "Find the video titled 'Q4 Roadmap' and update its privacy setting to public."
3. "Get the transcript for my most recent video and create a bulleted summary of the key action items discussed."

Tips & Limitations

• Permissions: Ensure your API key has the necessary scopes to perform delete or update actions if you plan to use those specific functions.
• Rate Limiting: The Loom API enforces rate limits; avoid high-frequency polling of video insights to prevent service disruption.
• Transcript Availability: Transcripts are generated by Loom upon upload. If a video is very new, the transcript might not be available immediately via the API.
• Data Security: Always store your LOOM_API_KEY in a secure environment variable. Do not hardcode it directly into scripts or share it in public repositories.