Second Brain

What is Second Brain?

Second Brain solves a persistent frustration for developers: every new AI chat starts from zero, forcing users to re-explain their tech stack, decisions, and preferences again and again. It runs as a Cloudflare Worker connected via Model Context Protocol, so any MCP-compatible client—Claude Desktop, Claude Code, ChatGPT, Cursor, Windsurf—shares one continuity layer instead of siloed, black-box memory. Beyond simple storage, it synthesizes patterns, resolves contradictions, and merges duplicate entries so the knowledge base stays clean over months, all while remaining fully self-hosted on the user's own free-tier Cloudflare account.

Key Features

• Universal MCP Connection

  One Worker URL plugs into Claude, ChatGPT, Cursor, Windsurf, and any other MCP-compatible client, sharing a single memory layer everywhere.

• AI Synthesis & Pattern Derivation

  Direct, synthesized answers to questions like past decisions, plus automatic surfacing of recurring themes never explicitly written down.

• Contradiction Detection & Smart Merge

  Semantic comparison flags conflicting memories and auto-resolves to the latest truth, while near-duplicate entries merge into one coherent record.

• Semantic Recall

  Meaning-based search retrieves relevant memories even when queries don't match stored wording exactly, with no manual tagging required.

• Web Dashboard

  Browse, filter, edit, and export all stored memories to JSON or Markdown through a full UI hosted directly at the Worker URL.

• Self-Hosted on Cloudflare

  Runs entirely on the user's own Cloudflare account using D1, Vectorize, and Workers AI—no vendor access, no credit card, no lock-in.

Use Cases

• Cross-Tool Development Continuity : Developers switching between Claude Code, Cursor, and ChatGPT keep the same architecture decisions and naming conventions available in every tool.
• Client & Project Tracking : Freelancers and consultants store client preferences, availability, and project history so the AI recalls status instantly in new sessions.
• Meeting and Note Capture : Teams pipe meeting notes or terminal output straight into memory via the CLI or Obsidian plugin for instant future recall.
• Mobile Thought Capture : iOS Shortcuts let users voice-dump ideas on the go, which sync directly into the AI's long-term memory.
• Long-Term Knowledge Maintenance : MemoryOS tracks decay and confidence scores across stored knowledge, flagging stale or unused entries before they mislead the AI.

FAQs