Organise Zoom summaries with AI
Save existing meeting summaries and repair uncertain terminology with a traceable glossary.
Organise Zoom summaries with AI when saved meeting notes contain mistranscribed domain terms or are awkward to retrieve. zoom-summary reads existing AI Companion summaries through authorised Zoom API access and saves local Markdown. It cannot create a missing summary or truthfully relay meeting content it never fetched.
Corrections distinguish high-confidence replacements from terms needing human review. The raw version remains available for repeatable edits. The guide also includes Korean translation and optional cloud-recording cleanup; deletion requires an agreed selection and explicit confirmation, with trash as the normal route rather than permanent removal.
What it does
- Retrieve summaries for a selected meeting or date range
- Save local text while preserving the raw original
- Apply confirmed glossary replacements consistently
- Leave uncertain names and terms for human review
- List recording sizes before deciding what to remove
- Hand selected decisions to a research log when requested
Install and use
- Use an authorised Zoom account with saved summaries, the required OAuth app and curl, jq, base64 and Python.
- Create the app yourself and configure its allowed scopes and local credentials securely.
- Retrieve a meeting summary and review uncertain terminology before accepting corrections.
- For recording cleanup, confirm the exact list before deletion; request permanent removal separately.
FAQ
Source
Made by Tae-Geun Kim. Original skill: zoom-summary, published on GitHub under the MIT licence.
View on GitHub