Granola Migration Deep DiveSAFE

Model-agnostic agent-skills platform with a harness-free canonical layer, verified adapters, and the ccpi package manager. Explore at tonsofskills.com.

The instruction file, verbatim from the audited commit — this is the text the model reads, and the surface the audit's instruction layer examines. Quoted here so you can judge it without cloning anything.

---
name: granola-migration-deep-dive
description: 'Migrate to Granola from Otter.ai, Fireflies, Fathom, tl;dv, or manual
  note-taking.

  Covers data export from source tools, parallel-run strategy, team transition,

  and historical data preservation.

  Trigger: "migrate to granola", "switch to granola", "granola from otter",

  "granola from fireflies", "replace meeting tool with granola".

  '
allowed-tools: Read, Write, Edit, Bash(python3:*)
version: 1.13.0
license: MIT
author: Jeremy Longshore <[email protected]>
tags:
- saas
- granola
- migration
compatibility: Designed for Claude Code
---
# Granola Migration Deep Dive

## Overview

Comprehensive guide for migrating to Granola from competing meeting note tools. Covers source-specific export procedures, historical data preservation, parallel-run strategy, team transition, and cutover execution. Granola's key differentiator — no bot joins meetings — means the migration also changes the user experience fundamentally.

## Prerequisites

- Access to source tool with export capability
- Granola workspace configured (see `granola-install-auth`)
- Migration timeline agreed with stakeholders
- Budget approved for Granola licenses

## Instructions

### Step 1 — Assess Migration Scope

```markdown
## Migration Assessment

Source tool: [Otter.ai / Fireflies / Fathom / tl;dv / Manual / Other]
Total meetings in source: [____]
Date range: [____] to [____]
Active users to migrate: [____]
Integrations to recreate: [Slack, Notion, CRM, etc.]
Historical data priority: [Archive all / Selective / Fresh start]
Target cutover date: [____]
Parallel run duration: [1 week / 2 weeks]
```

### Step 2 — Source-Specific Export

#### From Otter.ai

- **Export format:** TXT, SRT (subtitles), PDF
- **Bulk export:** Otter Pro/Business: Settings > Export > Download All
- **Limitations:** Free plan only exports individual notes
- **Key data:** Transcripts with timestamps, speaker labels, action items

#### From Fireflies.ai

- **Export format:** TXT, JSON, PDF, SRT
- **Bulk export:** Admin > Data Management > Export
- **Limitations:** Custom fields may not export
- **Key data:** Transcripts, AI summaries, custom topics

#### From Fathom

- **Export format:** Markdown, CSV, video clips
- **Bulk export:** Settings > Data > Export All
- **Limitations:** Video clips don't transfer
- **Key data:** Meeting summaries, action items, highlights

#### From tl;dv

- **Export format:** TXT, video recordings
- **Bulk export:** Settings > Data Export
- **Limitations:** AI highlights may not transfer
- **Key data:** Transcripts, timestamps, meeting recordings

#### From Manual Notes (Google Docs, Notion, Confluence)

- Already in accessible format
- No export needed — archive in place
- Focus on establishing the Granola workflow going forward

### Step 3 — Choose Migration Strategy

| Strategy | When to Use | Data Handling | Effort |
|----------|------------|--------------|--------|
| **Fresh Start** | <100 historical meetings, or meetings have low reference value | Archive source data externally, start fresh in Granola | Low |
| **Selective Migration** | 100-1000 meetings, some have ongoing reference value | Export key meetings (client calls, decisions, contracts) | Medium |
| **Full Archive** | Enterprise with compliance requirements, everything must be searchable | Export all data, archive in Notion/Drive/cloud storage | High |

**Recommended for most teams:** Fresh Start or Selective. Granola doesn't import historical data from other tools — there's no import feature. Historical data lives in your archive (Notion, Google Drive, local files).

### Step 4 — Archive Historical Data

For important historical meetings, archive before cutting over:

```python
#!/usr/bin/env python3
"""Organize exported meeting notes for archival."""
import os
from pathlib import Path

EXPORT_DIR = Path("~/Downloads/otter-export").expanduser()  # Adjust for your source
ARCHIVE_DIR = Path("~/Documents/meeting-archive").expanduser()

# Create organized archive structure
categories = {
    "client": ["client", "customer", "deal", "sales", "demo"],
    "engineering": ["sprint", "standup", "architecture", "review", "retro"],
    "product": ["product", "prd", "design", "feedback", "roadmap"],
    "leadership": ["all-hands", "board", "strategy", "planning"],
    "general": [],  # catch-all
}

ARCHIVE_DIR.mkdir(parents=True, exist_ok=True)
for cat in categories:
    (ARCHIVE_DIR / cat).mkdir(exist_ok=True)

for file in EXPORT_DIR.glob("*.txt"):
    filename_lower = file.name.lower()
    placed = False
    for cat, keywords in categories.items():
        if any(kw in filename_lower for kw in keywords):
            dest = ARCHIVE_DIR / cat / file.name
            file.rename(dest)
            placed = True
            break
    if not placed:
        (ARCHIVE_DIR / "general" / file.name).rename(file)

print(f"Archived to {ARCHIVE_DIR}")
```

Alternatively, upload the archive to Notion or Google Drive for team-wide searchability.

### Step 5 — Parallel Run (2 Weeks)

Run both tools simultaneously to build confidence:

**Week 1: Dual recording**

- Keep source tool active (bot still joins or captures)
- Enable Granola on all meetings (system audio capture)
- Compare output quality daily:

| Metric | Source Tool | Granola | Winner |
|--------|-----------|---------|--------|
| Transcription accuracy | ___% | ___% | |
| Action item detection | ___/total | ___/total | |
| Summary quality | ___/5 | ___/5 | |
| Processing time | ___ min | ___ min | |
| User experience (no bot) | N/A | Yes | Granola |

**Week 2: Granola primary**

- Keep source tool as backup only (disable auto-record if possible)
- All sharing and distribution via Granola integrations
- Team members report any quality issues

### Step 6 — Cutover Execution

**Cutover day checklist:**

- [ ] Final export from source tool (last day of data)
- [ ] Verify archive is complete and accessible
- [ ] Disable source tool recording/bot
- [ ] Remove sou

SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.

What the source does

Filesystem

none-observed

Network

none-observed

Shell

none-observed

Dependencies

pinned

Secrets in source

none-found

Findings (0)

No findings outside the package's declared scope.

Gates applied: no_behavioural_pass.