Granola Meeting Transcripts

**Judgment:** A solid, pragmatic skill for users who already rely on Granola and want local, scriptable access to their meeting history. It’s functionally strong for search and retrieval but constrained by platform (macOS) and setup complexity. **What it does well** - Syncs Granola meetings to a clear on-disk structure with metadata, transcript (raw + Markdown), API payload, and AI notes. - Provides concrete examples for listing recent meetings, searching by title or content, and filtering by date. - Leverages local files, so read/query performance should be fast and independent of Granola’s API once synced. **Key risks / limitations** - **Platform lock-in:** Auth path is macOS-specific; the documented setup assumes macOS and the Granola desktop app. - **Auth fragility:** Tokens expire in ~6 hours; if the user doesn’t periodically open the Granola app, sync may silently fail or lag. - **Operational overhead:** Requires manual installation (`pip install requests`), running a sync script, and configuring a cron job via `clawdbot_cron`. - **Data sensitivity:** Meeting transcripts and notes usually contain sensitive information; improper filesystem permissions, sharing, or downstream processing could expose private data. **Best-fit scenarios** - You are a macOS user already signed into Granola, comfortable with basic CLI/Python setup. - You need fast, repeated access to past meeting transcripts and summaries for search, reporting, or manual review. - You want a stable local mirror of Granola data to feed into other OpenClaw workflows, analytics, or custom automation. **Avoid / less ideal scenarios** - Multi-OS environments where macOS isn’t guaranteed, or where you can’t easily run the sync machine and rsync to others. - Users who aren’t comfortable managing cron jobs, local scripts, or secure handling of sensitive meeting data. - Situations requiring real-time or near-real-time data if sync schedules are infrequent or tokens often expire.