← Back to Home

Overview

SecondBrain Pro can connect to your own Cloudflare Workers backend, giving you full control over your data with global edge performance (<50ms latency). The backend is powered by shodh-cloudflare, an open-source distributed AI memory system.

Prerequisites

• A Cloudflare account (free tier is sufficient)
• Node.js 18+ installed
• Wrangler CLI: npm install -g wrangler
• Git installed

Step-by-Step Setup

1 Clone the Repository

git clone https://github.com/doobidoo/shodh-cloudflare.git
cd shodh-cloudflare

2 Create D1 Database

wrangler d1 create shodh-memory

Note the database_id from the output — you'll need it in the next step.

3 Create Vectorize Index

wrangler vectorize create shodh-vectors --dimensions=384 --metric=cosine

4 Configure wrangler.toml

cp worker/wrangler.toml.example worker/wrangler.toml

Edit worker/wrangler.toml and replace the database_id with your value from step 2.

5 Initialize Database Schema

cd worker
wrangler d1 execute shodh-memory --file=../schema.sql --remote

6 Set Your API Key

wrangler secret put API_KEY

Choose a strong API key — you'll enter this same key in the SecondBrain iOS app.

7 Deploy

npm install
npm run deploy

Your worker will be deployed to https://shodh-memory.YOUR-SUBDOMAIN.workers.dev

8 Verify

curl https://shodh-memory.YOUR-SUBDOMAIN.workers.dev/

You should see: {"message":"SHODH Memory API is running"}

Connect SecondBrain iOS App

1. Open SecondBrain Pro → Profile tab
2. Tap Not Configured under API Configuration
3. Enter your API Key (from step 6)
4. Enter your Base URL: https://shodh-memory.YOUR-SUBDOMAIN.workers.dev
5. Tap Save & Validate

What's Included

AI-Powered Features

• Smart Classification — Memories are auto-categorized (observation, decision, learning, etc.)
• Auto-Tagging — 2-4 contextual tags generated per memory
• Voice Correction — AI fixes dictation errors from Siri/Watch input
• Semantic Search — Find memories by meaning, not just keywords
• AI Summarization — Ask questions and get summaries across memories

Temporal Queries

Search with natural language time expressions:

• "What did I learn last week?"
• "Show me decisions from 3 days ago"
• "Memories from yesterday about the project"

Emotional Metadata

Each memory can track emotional context with valence, arousal, and emotion labels — helping you understand patterns in how you feel about different topics.

Costs

Next Steps

• Set up MCP Integration — Connect your memories to Claude Code, Cursor, or VS Code
• See all features
• GitHub Repository — Full source code, documentation, and troubleshooting