01One command setup
From empty folder to live MCP server
Scaffold files, install dependencies, create Cloudflare D1, migrate, deploy, and print connector instructions without hand-copying boilerplate.
Your AI context.
Connected everywhere.
Context Hub is your personal context layer and signal console. Sync across clients. Control your memory. Deploy to the edge.
>_
npx create-context-hubFree · Open source · MIT licensed★ Star on GitHub
Signal Console
Inspect, filter, and shape your context.
Context Layer
Store preferences, memories, and rules.
Cross-Client Sync
Real-time sync across all your tools.
Cloudflare Ready
Workers runtime. Global. Fast. Secure.
Claude
ChatGPT
Perplexity
Cursor
Claude Code
Codex
>_
Context Hub
Your context layer
Memories
Preferences
Rules
Signals
Edge Runtime
Global regions
Low Latency
Sub-50ms signals
Your Data
Encrypted. Yours.
Built for Builders
Open. Extensible.
36 weekly
Apr 29 – May 5
1.4K total
Since Mar 2026
Tired of re-explaining context to every AI client?One MCP context layer. Synced. Yours. Forever.
Trusted with context across
- Claude
- ChatGPT
- Perplexity
- Cursor
- Claude Code
- Codex
Features
Built for people whose AI work happens everywhere.
The shortest path from repeated prompting to durable, searchable context shared by every AI client you already use.
02Cross-client memory
Your context follows every AI client
Claude.ai, Claude Code, ChatGPT, Perplexity, Cursor, Windsurf, Zed, and custom MCP clients can read and write the same source of truth.
03Source-aware entries
Know where every memory came from
Context Hub tags memories and logs from the MCP initialize handshake, so entries can show whether they came from chatgpt, claude-code, or another client.
04Free cloud runtime
Cloudflare Workers and D1, not local daemons
The generated hub runs globally on Cloudflare's free tier with SQLite-backed storage, full-text search, and no Docker stack to babysit.
Use cases
Real workflows where shared context replaces re-prompting.
Pick a workflow, deploy in minutes, and let your AI clients reach the same source of truth.
01Sovereignty
Own your AI memory before a vendor wipes it for you
ChatGPT lost users' saved memories twice in 2025. Hosted AI memory is rented, not owned. Here's what changes when the rows live in your account, not theirs.
Read more →02Workflow
Sync Claude.ai memories into Claude Code
Save a product decision in the browser, then ask your terminal agent to use that same context while editing code.
Read more →03Infrastructure
Build a personal MCP memory server for $0/month
Deploy a durable AI context layer on Cloudflare Workers and D1 without paying for a hosted memory product.
Read more →04Multi-client
Share context between ChatGPT, Perplexity, and Cursor
Use Context Hub as the bridge when research, planning, and implementation happen in different AI tools.
Read more →Ship in one command
Stop re-explaining yourself
to every AI client.
One command. Every AI tool you use, finally on the same page.
>_
npx create-context-hub