AI Context Flow

What is AI Context Flow?

AI Context Flow is an OAuth-secured Model Context Protocol (MCP) server that gives any MCP-compatible AI client (Claude Code, Claude Desktop, Cursor, ChatGPT, and more) persistent, shared memory. Save, organize, and semantically search documents, notes, conversations, and files across memory buckets, so every AI tool you use works from the same context instead of starting from scratch each session.

It also comes with a companion browser extension, so you can capture and reuse your context directly on any website, not just inside your AI client.

How to use AI Context Flow?

AI Context Flow is a hosted remote server, so there is nothing to install or self-host. Point your MCP client at https://app.plurality.network/mcp and authenticate on first connect (a one-time browser OAuth login, or a Personal Access Token for headless agents).

• Claude Code: run claude mcp add --transport http plurality-memory https://app.plurality.network/mcp, then authenticate with /mcp.
• Claude Desktop: add it as a custom connector, or use the mcp-remote bridge in your config pointing at the URL above.
• Cursor and other streamable-HTTP clients: add the server using the Config block on this page (type: streamable-http, production URL).
• ChatGPT (paid plan): add it as a connector and authorize via OAuth.
• Headless agents (n8n, LangChain, CI): authenticate with a Personal Access Token sent as a Bearer token.

(Read docs for detailed guides for each platform.)

After the one-time login, your memory is available to that client automatically.

Key features of AI Context Flow

• Persistent, shared memory across all your AI clients, so Claude Code, Claude Desktop, Cursor, and ChatGPT all read and write the same context
• Semantic search across memory buckets with relevance scoring
• Organize context into memory buckets, and share buckets with other people
• Companion browser extension to capture and reuse context on any website, with a built-in sidebar chat agent (30+ models)
• 7 MCP tools covering bucket listing, semantic search, content read/write, and saving conversations
• Secure by default: OAuth 2.1 + PKCE for interactive clients, Personal Access Tokens for headless agents

Use cases of AI Context Flow

• Give Claude, Cursor, and ChatGPT a shared long-term memory that persists across sessions
• Semantic search across your stored documents, notes, and past conversations
• Save important chats straight into the right memory bucket for later recall
• Capture context from any web page with the browser extension and reuse it in your AI tools
• Collaborate by sharing memory buckets with teammates
• Power headless agents (n8n, LangChain, CI runners) with the same memory via PAT auth

FAQ from AI Context Flow

Which MCP clients are supported?

Claude Desktop, Claude Code (terminal and VS Code extension), ChatGPT (paid plan), Cursor, and any client that supports streamable HTTP and Dynamic Client Registration.

Is there a browser extension?

Yes. The AI Context Flow browser extension lets you capture and reuse context on any website, and includes a built-in sidebar chat agent with 30+ models that can answer from your stored memory. Use it alongside the MCP server to get the full product.

What authentication methods are available?

Interactive clients use OAuth 2.1 + PKCE (a one-time browser login). Headless agents and custom integrations use a Personal Access Token (PAT), managed from your dashboard.

Does AI Context Flow require a paid plan?

OAuth-based access is available on free plans. Personal Access Tokens and the ChatGPT connector require a paid plan (Pro, Max, Team, or Enterprise).

Do I need to install or host anything?

No. AI Context Flow is a hosted remote MCP server, so you just point your client at the URL and authenticate. The source is on GitHub if you want to inspect it, but self-hosting is not required to use the product.