granola

You can extract action items and attendees from meeting transcripts by applying this Skill to parse Granola notes. It processes raw meeting caches locally, identifying attendees, companies discussed, and TODOs with owners and deadlines, saving structured markdown to ~/.granola-archivist/output/.

The best way to analyze meeting notes locally for privacy is using a local processing tool that guarantees data never leaves your machine. This Skill extracts intelligence from Granola caches directly on your device, ensuring privacy-first handling of meeting transcripts and actionable insights.

You do not strictly need an OpenAI API key to extract meeting intelligence, but it is an optional prerequisite for enabling AI-powered extraction. Without it, the Skill can still pull meeting metadata and attendees from Granola caches, saving structured markdown locally.

To convert raw meeting transcripts into structured markdown, tell the Skill to extract your Granola meetings. It identifies attendees, companies, and key insights, generating structured markdown files that highlight TODOs and actionable items, saved directly to ~/.granola-archivist/output/.

Granola meeting extraction works without an internet connection for local data processing, as all metadata and attendee extraction happens on your machine. However, if you utilize the optional OpenAI API key for AI intelligence extraction, an internet connection is required.

Extracted meeting notes and insights are saved locally to the ~/.granola-archivist/output/ directory. The Skill generates structured markdown files in this folder, containing attendees, companies discussed, TODOs, and key insights parsed from your Granola caches.