instagram-import

Instagram Import

Created by @alanshurafa

Reviewed and merged by the Open Brain maintainer team — thank you for building the future of AI memory!

Import your Instagram data — DMs, comments, and post captions — into Open Brain.

What It Does

Parses Instagram's data export and imports three types of content as searchable thoughts:

• Messages — DM conversations (minimum 3 messages per conversation)
• Comments — Your comments on posts, batched together
• Posts — Your post captions, batched together

Handles Meta's double-encoded UTF-8 text (latin1 → UTF-8 conversion).

Prerequisites

• Working Open Brain setup (guide)
• Instagram data export — download from Instagram Settings
• Node.js 18+ installed
• OpenRouter API key for embedding generation

Credential Tracker

INSTAGRAM IMPORT -- CREDENTIAL TRACKER
--------------------------------------

FROM YOUR OPEN BRAIN SETUP
  Supabase URL:          ____________
  Service Role Key:      ____________

FROM OPENROUTER
  API Key:               ____________

--------------------------------------

Steps

1. Request your Instagram data:

   • Go to Instagram → Settings → Accounts Center → Your information and permissions → Download your information
   • Select JSON format
   • Download and extract the archive
   • Look for the your_instagram_activity/ folder

2. Copy this recipe folder and install dependencies:

   cd instagram-import
   npm install

3. Create .env with your credentials (see .env.example):

   SUPABASE_URL=https://your-project.supabase.co
   SUPABASE_SERVICE_ROLE_KEY=your-service-role-key
   OPENROUTER_API_KEY=sk-or-v1-your-key

4. Preview what will be imported (dry run):

   node import-instagram.mjs /path/to/instagram-export --dry-run

5. Import specific types only (optional):

   node import-instagram.mjs /path/to/instagram-export --types messages
   node import-instagram.mjs /path/to/instagram-export --types comments,posts

6. Run the full import:

   node import-instagram.mjs /path/to/instagram-export

Expected Outcome

After running the import:

• DM conversations become thoughts tagged with source_type: instagram_import
• Long conversations are capped at 200 messages per thought
• Comments and captions are batched (50 comments or 30 captions per thought)
• All content with sensitivity_tier: personal
• Running search_thoughts { query: "that restaurant recommendation" } finds relevant DMs

Scale reference: Tested with 502 Instagram items imported successfully.

Troubleshooting

Issue: "Could not find your_instagram_activity directory" The export structure varies by download method. Look inside your extracted archive for a folder named your_instagram_activity. If it's nested deeper, point the script at the parent folder.

Issue: Garbled text (wrong characters) Meta exports encode text as latin1-interpreted UTF-8. The script fixes this automatically with fixMetaEncoding(). If text still looks wrong, the file may use a different encoding.

Issue: No messages found DMs are in your_instagram_activity/messages/inbox/. Each conversation is in its own folder with message_1.json, message_2.json, etc. Check that this structure exists in your export.