vt-c-meeting-minutes¶

This skill should be used when creating meeting minutes from transcripts, screenshots (Miro boards, whiteboards), or other meeting documentation. It generates structured Markdown meeting minutes with versioning (0.1 → 0.2 → ... → 1.0) and automatically produces a branded PDF via the document-converter-branded skill.

Plugin: core-standards
Category: Other
Command: /vt-c-meeting-minutes

Meeting Minutes Skill¶

Transform meeting transcripts and visual inputs (Miro boards, whiteboards, etc.) into structured, versioned meeting minutes in Markdown format.

Invocation¶

# Interactive mode - prompts for all options
/vt-c-meeting-minutes

# With transcript file
/vt-c-meeting-minutes /path/to/transcript.txt

# With screenshot
/vt-c-meeting-minutes /path/to/miro-screenshot.png

Execution Instructions¶

When this skill is invoked, follow these steps in order:

Step 1: Initial Setup¶

Use AskUserQuestion to gather required information:

Language: German or English?
Product Branding: VisiTrans, VisiMatch, VisiFair, or VisiArea?
Output Path: Where to save the meeting minutes?
Meeting Topic: Brief description for the filename

Step 2: Gather Input¶

Accept one or more of: - Transcript text: Pasted directly or file path - Screenshot paths: Miro boards, whiteboards, sticky notes - Audio file references: If mentioned, note that transcription is needed first

For screenshots, use vision capabilities to: - Extract text from sticky notes and cards - Identify groupings and relationships - Note any visual hierarchy or categorization

Step 3: Extract Meeting Information¶

From the input, identify:

Participants: Names and organizations
Extract all names mentioned in transcript/screenshots

Always present found participants to user for validation:

"I found the following participants in the transcript:

Organization 1:
- Name A
- Name B

Organization 2:
- Name C

Is anyone missing? (Silent participants may not appear in transcripts)
Are the organization groupings correct?"

If no participants found or unclear, use AskUserQuestion:

"Please provide the meeting participants grouped by organization.
Format: Organization1: Name1, Name2 | Organization2: Name3, Name4"

Meeting Date: Extract from transcript or ask user
Meeting Purpose: The main topic or goal
Key Findings: Discussion points, decisions, requirements
For each finding, extract:
- Title/Topic
- Requirement or decision
- Background/context
- Implementation details or next steps
Action Items: Tasks with owners and priorities
Group by phase or priority level
Open Items: Unresolved questions for follow-up
Next Steps: Immediate, this week, next meeting

Step 4: Generate Meeting Minutes¶

Select Template based on language:
German: templates/protokoll-de.md
English: templates/minutes-en.md
Set Logo Path based on product (resolve BRAND_ASSETS_ROOT per CLAUDE.md Path Resolution):
```
BRAND_ASSETS_ROOT/logos/[Product].jpg
```

Product	Logo File
VisiTrans	VisiTrans.jpg
VisiMatch	VisiMatch.jpg
VisiFair	VisiFair.jpg
VisiArea	VisiArea.jpg

Create File with naming convention (all lowercase, spaces as separators):
German: yyyy-mm-dd protokoll [meeting-topic] v0.1.md
English: yyyy-mm-dd minutes [meeting-topic] v0.1.md
Example: 2026-02-04 protokoll consigneepage review 2 v0.1.md
Fill Template with extracted information:
Replace all {{PLACEHOLDER}} values
Set Version to 0.1
Set Status to Entwurf (German) or Draft (English)
Set Author from user context or ask
Write File to specified output path

Step 5: Generate PDF¶

Always generate a PDF alongside the Markdown file:

Invoke document-converter-branded skill:
```
python scripts/convert_to_pdf.py \
  --input [minutes-file.md] \
  --output [minutes-file.pdf] \
  --product [visitrans|visimatch|visifair|visiarea]
```
The PDF uses the same filename as the Markdown file but with .pdf extension. Example: 2026-02-04 protokoll consigneepage review 2 v0.1.pdf
Validate PDF was created successfully

Step 6: Present Output¶

Show user: - File paths of both Markdown and PDF - Summary of extracted information - Any items that need clarification

Ask if any corrections are needed. If corrections are made, regenerate the PDF.

Step 7: Version Management (When Updating)¶

When updating existing minutes:

Read current version from filename or metadata
Increment version following rules in references/versioning.md:
```
0.1 → 0.2 → ... → 0.9 → 0.10 → 0.11 → ...
```
For final approval: Change to 1.0, update status to Final/Freigegeben
Update version history table with change description

Consider using the doc-versioning skill to: - Archive previous version to 99-alt/ folder - Handle version number increment automatically

Step 8: Version Update with PDF¶

When updating an existing version, always regenerate the PDF for the new version as well. Both the Markdown and PDF files use the same naming convention with the updated version number.

Template Structure¶

The meeting minutes follow this structure:

Header: Product logo, document title
Metadata: Version, date, status, author
Version History: Table tracking all versions
Meeting Overview: Date, participants (by organization), purpose
Key Findings: Numbered sections with requirements, background, implementation
Action Items: Phased to-do list with checkboxes
Open Items: Unresolved questions for follow-up
Next Steps: Immediate, this week, next meeting actions
Summary: Key takeaways and goal

Versioning Rules¶

See references/versioning.md for complete rules.

Quick Reference: - Draft versions: 0.1 → 0.2 → ... → 0.9 → 0.10 → 0.11 → ... - Final version: 1.0 (when approved) - After 0.9 comes 0.10, NOT 1.0

Integration with Other Skills¶

Skill	Usage
`visitrans_cd`	Logo files from `assets/logos/`
`document-converter-branded`	PDF generation with product branding
`doc-versioning`	Archive old versions, increment version numbers

Example Output¶

German¶

2026-02-04 protokoll consigneepage review 2 v0.1.md
2026-02-04 protokoll consigneepage review 2 v0.1.pdf

English¶

2026-02-04 minutes consigneepage review 2 v0.1.md
2026-02-04 minutes consigneepage review 2 v0.1.pdf

Tips for Best Results¶

Provide clear transcripts: Well-formatted transcripts yield better extraction
Include participant context: Mention organizations when listing participants
Mark decisions clearly: Use phrases like "decided", "agreed", "action item"
Separate topics: Use clear topic transitions in transcripts
Note open questions: Explicitly mark unresolved items
Review participant list: Silent attendees won't appear in transcripts - always check for missing names

Error Handling¶

Missing Participants¶

Always validate extracted participants with user - silent attendees won't appear in transcripts.

If participants cannot be extracted at all, prompt user:

"Please list the meeting participants:
Organization 1: Name1, Name2
Organization 2: Name3, Name4"

If some participants were found, show them and ask:

"I found these participants: [list]. Is anyone missing?"

Missing Date¶

If date is unclear, use current date and confirm with user.

Unclear Decisions¶

Mark unclear items as "CLARIFICATION NEEDED" and highlight for user review.

Open Brain Capture (Optional)¶

After producing the minutes document, if the capture_thought MCP tool is available, capture the meeting summary to Open Brain.

When to capture: After every finalized minutes document.

How: 1. Check if capture_thought tool is available. If not: skip silently. 2. If the minutes content is empty: skip capture. 3. Call capture_thought with:

thought: "Meeting: {topic} on {date} with {participants}. Decisions: {list of decisions}. Actions: {list of action items with owners}."

4. On timeout or error: log debug message and continue. Never fail the skill.

This step is always last — it never interrupts the meeting-minutes workflow.