v1.1.0 - Interactive Prompting

Interactive Prompting Enhancement

New in v1.1.0: The skill now guides agents to ask clarifying questions before starting exports, ensuring better user experience and correct parameters.

New Features

Interactive Prompting

• Asks for date range (last 7/30 days, custom range)
• Asks for export location (defaults to ./granola_exports/)
• Asks for meeting selection (all or specific)

Fixed Requirements (prevents confusion)

• Prominent section explaining non-negotiable format standards
• Always markdown .md files (never JSON, TXT, etc.)
• Always summary + full transcript (both required)
• Always one file per meeting (never combined/split)

Smart Defaults

• Last 30 days if no date range specified
• ./granola_exports/ if no location specified
• All meetings if no selection specified

Example Interaction

User: "Export my Granola meetings"

Agent: "I'll help you export your Granola meetings. Let me gather a few details:

1. Which time period?
   - Last 7 days
   - Last 30 days (recommended)
   - Custom date range

2. Where should I save the files?
   - Default: ./granola_exports/
   - Or specify a path

[After fetching meetings]
Found 8 meetings from Aug-Dec 2025
3. Export all 8 or select specific ones?"

Testing

Validated with TDD methodology:

• RED phase: Tested baseline behavior
• GREEN phase: Added prompting guidance
• REFACTOR phase: Clarified fixed requirements

Upgrading from v1.0.0

Simply update your skill file - no breaking changes. The skill now provides better UX through interactive prompting while maintaining the same quality standards.

---

Full Changelog: https://github.com/jtklinger/granola-transcripts-skill/blob/master/CHANGELOG.md

v1.0.0 - Initial Release

Exporting Granola Transcripts - Claude Code Skill

First stable release of the Granola transcript export skill with verified completeness checking and proper error handling.

Features

✅ Binary Success Criteria - ALL transcripts complete or FAIL (no "partial success")
✅ Verification Checklist - Character count, natural endings, cutoff detection
✅ Red Flags System - Prevents success bias with prohibited phrase detection
✅ Error Handling - Automatic retries with exponential backoff
✅ Rate Limit Support - 60s delays between requests
✅ Single File Format - Combined markdown with summary + full transcript

Testing Methodology

Developed using TDD for documentation:

• 5 baseline tests documented default agent behavior without skill
• 3 GREEN/REFACTOR tests verified skill effectiveness
• Success scenario validated proper success reporting
• Failure scenarios confirmed proper error handling

Known Limitations

• Granola API rate limits may block large exports
• Rate limit timing may vary by account tier

Installation

# Copy to your Claude Code skills directory
cp -r exporting-granola-transcripts ~/.claude/skills/

See README.md for full documentation.

Credits

Developed using the writing-skills TDD methodology for Claude Code skills.