Meeting Minutes Assistant
Paste the following prompt into your AI chat to install this skill:
Follow https://skillhub.cn/install/skillhub.md to install @user_94b6ee61/meeting-render.
About this skill
Problem
Meeting Minutes Assistant works on meeting text that has already been transcribed and roughly organized upstream. It fixes unstable template rendering: inconsistent issue splitting, arbitrary conclusions, mixed problem types, and uncertainty around owners and product terms.
How It Works
It generates paste-ready Markdown according to templates/sample.md:
- Issue recognition: split by explicit headings; if absent, create a 10–20 character short title.
- Review type: use online review when the text says “online review / async review / comment-section review / online review”; otherwise default to meeting review.
- Conclusion: default to “Approved, no re-review needed”; switch to “Not approved, re-review required” only when explicit re-review semantics appear.
- Issue classification: emit 【Clarification】, 【Suggestion】, 【Question】, or 【Defect】; use Defect only when repair, error, or risk can be clearly inferred.
- Two-stage confirmation: draft first, then ask one item at a time to verify proposer, owner, details, or terminology before final output.
Boundaries
It does not perform speech recognition, raw note cleanup, or Jira status backfill; the Processing Result column remains blank. If terms are missing from the glossary, suspected transcription errors, or ownership is unclear, it asks the minimum necessary questions. Corrections are written as wrong term -> correct term and only persisted when the user explicitly confirms.
Use Cases
- Turn transcribed review notes into a fixed template with split issues and conclusions.
- Classify clarifications, suggestions, questions, and defects, then confirm owners.
- Verify suspected transcription terms before finalizing the minutes output.
- Distinguish online review from meeting review blocks based on explicit wording.
Best For
- Product managers organizing review minutes who need standardized templates from transcribed text.
- Project assistants tracking review outcomes who need issue types and owners confirmed.
- Office efficiency teams maintaining glossaries who need transcription errors corrected and terms stored.
- Engineers using upstream transcription drafts who need multi-topic minutes rendered as paste-ready Markdown.
Related Skills
Batch-import team weekly reports, aggregate progress, plans, issues, and support by project, flag delivery or resource risks, and generate structured department reports.
Generates Kingdee Cloud ERP startup, implementation, and acceptance documents from Word templates with delivery guidance.
Read PDF, Word, Excel, PowerPoint, HTML, and text documents, list archive files, and extract text or JSON for AI analysis.
Extract decisions and action items from meeting notes, emails, or chat logs, then track owners, deadlines, and follow-ups with lightweight reminders and no direct tool integrations.