A personal memory CLI. Tell it things in plain language; ask it later.
you › my friend told me there are sick jackets in Islamabad
bot › got it, I'll remember that.
you › where did I hear about jackets?
bot › Your friend mentioned there are some great jackets in Islamabad.
1. Install dependencies
pip install -r requirements.txt2. Create your .env file
cp .env.example .envOpen .env and fill in:
SUPABASE_URLandSUPABASE_KEY— from your Supabase project settings.GEMINI_API_KEY— from Google AI Studio.CHAT_MODELandEMBED_MODEL— confirm current model IDs from the Gemini docs before filling these in.
3. Set up the database
Run supabase/schema.sql in the Supabase SQL editor of your project.
4. Run
python -m recallOn the Gemini free tier, prompts may be used by Google for model training. Do not store sensitive personal data if this is a concern.
| Command | Purpose |
|---|---|
/add <text> |
Force-store text as a memory |
/ask <text> |
Force-treat text as a question |
/search <text> |
Show raw matching memories (no AI answer) |
/list |
Show all stored memories with their ids |
/forget <id> |
Delete a memory by id |
/count |
How many memories are stored |
/help |
Show help |
/quit |
Exit |
Or just type naturally — the app detects whether you're storing or asking (Phase 4+).