Getting Started

Set up Bindly in three steps. Give your AI a persistent knowledge layer that works across every conversation.

Step 1

Sign up at bind.ly/app with your email or Google account. A personal Space is created automatically — a private workspace for you and your LLM.

Step 2

Add Bindly to your MCP client (Claude Desktop, Claude Code, Cursor, or any MCP-compatible client):

{
  "mcpServers": {
    "bindly": {
      "url": "https://mcp.bind.ly/mcp"
    }
  }
}

On first use, you will be prompted to authenticate via OAuth. See the full MCP Connection Guide for details.

Step 3

Tell your AI something like:

Key Concepts

Bindly organizes knowledge into a few core concepts:

Binding — A knowledge container (article, note, conversation extract). Each update creates a new immutable Version.
Version — An immutable snapshot of a Binding's content. Nothing is ever lost.
Space — A workspace that owns Bindings and Sets. Types: personal (private), team (private/public toggle), open (community, permanently public).
Set — A curated collection of specific Versions, like a playlist for knowledge. Use it to load focused context within a token budget.

Once connected, your LLM has access to 29 MCP tools for:

Saving — mcp_create_binding stores content with summaries, key points, and entities
Searching — mcp_search finds knowledge by semantic or keyword search
Retrieving — mcp_get_binding with tiered content delivery (Tier 1/1.5/2) for token-efficient access
Organizing — mcp_create_set and mcp_get_set_context for context assembly within a token budget
Sharing — mcp_create_share generates temporary share links
Commenting — mcp_add_comment adds notes to Bindings or Versions

See the full MCP Tools Reference for details on all 29 tools.