# Knot

Knot shares project context between connected AI clients. Use it to recover work or carry decisions across agent sessions.

- MCP server · `type=mcp` · Claude Code, Codex, Cursor, GitHub Copilot, Windsurf (Devin Desktop), Cline, Zed, Gemini CLI
- Author: ConnectKnot
- License: not stated · policy=link-only · description only; the file stays with the author
- Source: https://github.com/ConnectKnot/knot-mcp
- Updated: 2026-10-27
- Security check: automated, 2026-11-16 (text) — no signs of malicious code found.
- HTML: https://affpapa.org/en/nevibecode/mcp/knot

## What it does

Knot gives connected AI clients access to shared project memory. Agents can retrieve current facts and earlier decisions with `get_relevant_context`, or resume work from another AI session with `get_context_recovery`. New project information can be stored through `propose_context` or `batch_propose_context`.

The server also supports structured task handoffs between agents and switching between personal and team spaces. Add it to Claude Code with the command below. On the first connection, a browser opens for sign-in and access approval using OAuth 2.1 with PKCE, so completing that authorization step is required.

## Requirements
- Remote server: `https://api.connectknot.com/mcp`

## Install

### Claude Code (native platform)

**Platform CLI: remote server**
```bash
claude mcp add --transport http knot https://api.connectknot.com/mcp
```

### Codex

**Platform CLI: remote server**
```bash
codex mcp add knot --url https://api.connectknot.com/mcp
```

### Cursor

**Config file**
File `.cursor/mcp.json`:
```json
{
    "mcpServers": {
        "knot": {
            "url": "https://api.connectknot.com/mcp"
        }
    }
}
```

### GitHub Copilot

**Config file**
File `.vscode/mcp.json`:
```json
{
    "servers": {
        "knot": {
            "type": "http",
            "url": "https://api.connectknot.com/mcp"
        }
    }
}
```

### Gemini CLI

**Platform CLI: remote server**
```bash
gemini mcp add -t http knot https://api.connectknot.com/mcp
```

## When to use it
- recover project context after switching AI agents
- retrieve saved project decisions and constraints
- save new facts and goals to shared memory
- hand a structured task to another connected agent

## Example prompts
- Use `get_context_recovery` to restore the context from my previous AI session.
- Find the current project decisions and constraints with `get_relevant_context`.
- Save this project decision with `propose_context`.

## Similar
- [ClariLayer](https://affpapa.org/en/nevibecode/mcp/clarilayer.md) — MCP server · ★ 68 repo stars
- [OpenMyMemo](https://affpapa.org/en/nevibecode/mcp/openmymemo.md) — MCP server · ★ 31 repo stars
- [orca-cli](https://affpapa.org/en/nevibecode/skills/orca-cli.md) — skill · 363100 installs
- [knot-theory-educator](https://affpapa.org/en/nevibecode/skills/knot-theory-educator.md) — skill · 182 installs
- [takeover](https://affpapa.org/en/nevibecode/skills/takeover.md) — skill · 1200 installs
- [pulumi-neo-handoff](https://affpapa.org/en/nevibecode/skills/pulumi-neo-handoff.md) — skill · 463 installs

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md
