⚡ Fireflies.ai to Obsidian Sync (All Meetings Edition)

A robust, state-based (Delta Sync) automation designed to extract transcripts and Artificial Intelligence metadata from Fireflies.ai and inject them directly into Obsidian as structured Markdown notes.

🤝 Credits & Architecture

This project is a deeply refactored architectural fork, born from the excellent foundation built by Ritwik-28. The original repository served as a great starting point for connecting to the Fireflies API.

However, to scale the script's reliability to a production level, fundamental architectural changes were implemented in this version:

1. From Time to State (Delta Sync): The original version relied on the current date to fetch meetings. If the host machine was offline for a weekend, meetings from that period were lost forever. This version implements a local "Ledger" (synced_ledger.json). The script asks the API for the latest meetings and cross-references them with the local Ledger. What isn't on the disk, it downloads. No meeting is lost, regardless of downtime.
2. AI Extraction (GraphQL): The request was expanded to pull not just raw transcripts, but the summary blocks (Overview, Action Items, and Shorthands) generated by Fireflies' built-in AI.
3. Local Timezone Fix (BRT / UTC-3): Corrected the handling of Epoch timestamps received from the API, which previously defaulted to standard UTC, pushing late-night meetings to the next calendar day.

⚙️ Core Features

• Silent Execution: Designed to run entirely in the background via Systemd without human intervention.
• Duplicate Prevention: The local Ledger strictly prevents Obsidian from creating duplicate files.
• YAML Frontmatter: Natively prepares the generated notes for Obsidian's Dataview plugin.
• Secure Isolation: Code structured to run inside Python virtual environments (venv), making it ideal for immutable systems like Fedora Atomic/Silverblue.

🛠️ Installation (Linux / Systemd Focus)

The following guide assumes deployment on a modern Linux environment (utilizing venv to avoid polluting the root system).

1. Cloning & Virtual Environment

git clone https://github.com/marcosaugustoldo/fireflies-obsidian-sync-all.git
cd fireflies-obsidian-sync-all
python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt

2. Credentials Configuration

Create your environment variables file:

cp .env.example .env
nano .env

Insert your API key and the absolute path to your Obsidian vault:

FF_API_KEY="your_fireflies_api_key_here"
OBSIDIAN_VAULT_PATH="/absolute/path/to/your/vault"

3. The First Run (Ledger Creation)

The initial execution will download all recent meetings and create the state file synced_ledger.json.

/path/to/your/venv/bin/python sync_fireflies.py

🕒 Automation with Systemd On Linux (Cron Replacement)

To ensure continuous, silent execution, create a Systemd user service.

1. Create the Service (~/.config/systemd/user/fireflies-sync.service):

[Unit]
Description=Fireflies.ai to Obsidian Synchronization

[Service]
Type=oneshot
WorkingDirectory=/absolute/path/to/the/script/folder
ExecStart=/absolute/path/to/your/venv/bin/python sync_fireflies.py

2. Create the Timer (~/.config/systemd/user/fireflies-sync.timer):

[Unit]
Description=Fireflies Sync Timer

[Timer]
OnCalendar=*-*-* 08..22:00:00
Persistent=true

[Install]
WantedBy=timers.target

3. Enable the Automation:

systemctl --user daemon-reload
systemctl --user enable --now fireflies-sync.timer

Your data pipeline is now fault-tolerant and will operate passively in the background.