Meeting Insights Analyzer

Intelligent meeting analysis toolkit that processes meeting transcripts, extracts action items, identifies key decisions, generates summaries, and tracks follow-ups across multiple meetings.

When to Use This Skill

Choose Meeting Insights Analyzer when:

• Processing recorded meeting transcripts for key takeaways
• Extracting action items and assigning owners from meeting notes
• Generating concise meeting summaries for absent stakeholders
• Tracking decisions and commitments across multiple meetings
• Analyzing meeting patterns for team productivity improvements

Consider alternatives when:

• Need real-time transcription — use Otter.ai, Fireflies, or Zoom transcription
• Need meeting scheduling — use calendar tools
• Need video conferencing — use meeting platforms directly

Quick Start

Copy
# Activate meeting analyzer
claude skill activate expert-meeting-insights-analyzer-kit

# Analyze a transcript
claude "Analyze this meeting transcript and extract action items: [paste transcript]"

# Generate summary
claude "Generate a concise summary of the sprint planning meeting transcript"

Example Analysis Output

Copy
## Meeting Analysis: Sprint 24 Planning
**Date:** 2024-03-15 | **Duration:** 45 min | **Attendees:** 6

### Summary
The team finalized Sprint 24 scope, prioritizing the checkout redesign
and payment integration. Two stories were deprioritized due to API
dependency on the partner team. The QA lead raised concerns about
test coverage on the payment flow.

### Key Decisions
1. Checkout redesign will use the new component library (decided by Sarah)
2. Payment integration blocked until partner API v2 ships (estimated March 22)
3. Sprint demo moved to Thursday due to company all-hands conflict

### Action Items
| # | Action | Owner | Due Date | Priority |
|---|--------|-------|----------|----------|
| 1 | Create checkout redesign PR with new components | Alex | Mar 19 | High |
| 2 | Write integration tests for payment flow | Jamie | Mar 20 | High |
| 3 | Follow up with partner team on API v2 timeline | Sarah | Mar 18 | Medium |
| 4 | Update sprint board with revised priorities | Mike | Mar 15 | Low |
| 5 | Share test coverage report with team | Jamie | Mar 19 | Medium |

### Risks & Blockers
- **Blocker:** Partner API v2 not yet available, blocking 2 stories
- **Risk:** QA test coverage may delay sprint completion
- **Risk:** New component library has known issues with Safari

### Follow-up Topics for Next Meeting
- Partner API v2 availability status
- Test coverage progress report
- Safari compatibility fix timeline

Core Concepts

Analysis Components

Component	Description	Output
Summary Generation	Concise overview of meeting purpose and outcomes	3-5 sentence summary
Decision Extraction	Identify decisions made, including who decided	Decision log
Action Items	Extract tasks with owners, deadlines, priorities	Action item table
Risk Identification	Surface concerns, blockers, and dependencies	Risk register
Sentiment Analysis	Detect areas of agreement, disagreement, or concern	Sentiment map
Topic Segmentation	Break transcript into distinct discussion topics	Topic timeline
Follow-up Tracking	Connect action items to previous meeting commitments	Follow-up report

Meeting Analysis Patterns

Meeting Type	Focus Analysis On	Key Outputs
Sprint Planning	Story commitments, capacity, dependencies	Sprint backlog, blockers
Stand-up	Progress, blockers, commitments	Updated task status
Retrospective	What worked, what didn't, improvements	Action items for process
Design Review	Decisions, trade-offs, next steps	Design decisions log
1:1	Goals, feedback, career development	Action items, notes
All-hands	Announcements, Q&A topics, morale	Key announcements summary

Copy
interface MeetingAnalysis {
  metadata: {
    title: string;
    date: string;
    duration: number;      // minutes
    attendees: string[];
    meetingType: string;
  };
  summary: string;
  topics: Topic[];
  decisions: Decision[];
  actionItems: ActionItem[];
  risks: Risk[];
  followUps: FollowUp[];
  sentiment: {
    overall: 'positive' | 'neutral' | 'negative';
    highlights: string[];
    concerns: string[];
  };
}

interface ActionItem {
  description: string;
  owner: string;
  dueDate: string;
  priority: 'high' | 'medium' | 'low';
  status: 'new' | 'in_progress' | 'completed' | 'blocked';
  context: string;  // Quote from transcript
}

Configuration

Parameter	Description	Default
meeting_type	Type of meeting for context-aware analysis	general
extract_actions	Extract action items with owners	true
extract_decisions	Extract key decisions	true
summary_length	Summary length: brief, standard, detailed	standard
track_followups	Cross-reference with previous meetings	false
output_format	Output: markdown, json, notion, jira	markdown
language	Transcript language	en

Best Practices

1. Process transcripts within 24 hours of the meeting — Context and memory fade quickly. Analyze transcripts the same day while attendees can still verify action items and correct misinterpretations of decisions.

2. Always attribute decisions and action items to specific people — Unowned action items don't get done. If the transcript doesn't clearly identify an owner, flag the item as "Owner: TBD" and follow up immediately rather than assigning arbitrarily.

3. Distinguish between decisions and discussions — Not every topic discussed results in a decision. Mark statements as decisions only when there's explicit agreement or authority exercised. Ongoing discussions should be tracked as "Open Items" for the next meeting.

4. Track action items across meetings for accountability — Maintain a running action item list that carries forward uncompleted items. Each meeting analysis should start by reviewing the status of previous action items before adding new ones.

5. Include relevant transcript quotes as evidence — When extracting decisions or action items, include the relevant transcript snippet. This provides context, prevents misinterpretation, and creates an auditable trail from summary back to source.

Common Issues

Transcription quality is poor, making analysis unreliable. Low-quality audio produces garbled transcripts. Pre-process by correcting obvious transcription errors for key terms (proper names, technical terms, project names). Create a glossary of domain-specific terms that transcription often mangles.

Action items are implicit rather than explicitly stated in meetings. Many commitments are implied through discussion context rather than stated as "I will do X by Y date." Train the analyzer to recognize implicit commitments like "I'll look into that," "Let me check," and "We should probably..." and flag them as potential action items needing confirmation.

Meeting summaries miss the nuance of complex discussions. Long technical debates don't compress well into bullet points. For complex topics, create a "Discussion Summary" section that captures the arguments for and against each option, then a separate "Decision" section for the final outcome. This preserves the reasoning behind decisions.