Metadata-Version: 2.4
Name: mcp-server-relay
Version: 1.0.0
Summary: MCP relay server for Claude Desktop <-> Claude Code communication
Project-URL: Homepage, https://github.com/mhcoen/mcp-relay
Project-URL: Repository, https://github.com/mhcoen/mcp-relay
Author-email: Michael Coen <mhcoen@gmail.com>
License-Expression: MIT
License-File: LICENSE
Keywords: anthropic,claude,mcp,relay
Classifier: Development Status :: 4 - Beta
Classifier: Intended Audience :: Developers
Classifier: License :: OSI Approved :: MIT License
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.9
Classifier: Programming Language :: Python :: 3.10
Classifier: Programming Language :: Python :: 3.11
Classifier: Programming Language :: Python :: 3.12
Requires-Python: >=3.9
Requires-Dist: mcp
Description-Content-Type: text/markdown

# Relay

A wonderfully simple tool that moves information—files, code, data, comments, directions, etc.—between Claude Desktop and Claude Code while keeping their main context windows separate.

Both clients connect to a shared buffer via MCP. Just say "ask Desktop" or "send this to Code"—the models handle the rest. Models will also automatically send messages if they determine they need information the other model has. System notifications alert you when something's waiting on the other side.

**Why?** Desktop and Code have different strengths. Desktop is better for conversation—planning, brainstorming, reviewing, iterating on prose. Code is better for execution—editing files, running commands, working through errors. But they don't share context. If you draft something in Desktop and want Code to implement it, or you want Desktop's opinion on code you're writing, you're copy-pasting between apps.

Relay connects them without requiring you to manually specify questions and cut & paste between windows. This is basic functionality that should have been integrated by Anthropic.

**You're always in control.** Messages sit in the buffer until you explicitly fetch them. The models can't pull from the relay on their own—you have to type `relay` or `/relay`. This means there's no runaway inter-LLM conversation burning through your tokens. You see a notification, you decide whether to check it.

## Quick Example

```
[In Desktop]
You:     Send this error to Code: "TypeError: cannot unpack non-iterable NoneType"
Desktop: [sends via relay]

[notification pops up in Code's terminal]

[In Code]
You:     /relay
Code:    Got it. That's from line 47 in parser.py—the regex isn't matching.
         [fixes the bug]
```

## Extended Example

```
[In Code]
You:     My classifier's stuck at 72% accuracy. Send the
         training config and last epoch's metrics to Desktop.
Code:    [sends config.yaml and metrics via relay]

[notification pops up in Desktop]

[In Desktop]
You:     relay
Desktop: [fetches config and metrics]
         The loss curve suggests you're underfitting. Learning
         rate 1e-5 is conservative for a model this size...

You:     What would you try?
Desktop: Bump to 3e-5, add warmup steps, and your batch size
         of 8 is small—if memory allows, try 16. Also the
         class imbalance is brutal: 80/15/5 split. Consider
         weighted loss or oversampling the minority classes.

You:     Specify the changes.
Desktop: [sends updated config via relay]

[System notification appears that a message has been sent to Code]

[In Code]
You:     /relay
Code:    Got it. Updating config.yaml and adding class weights
         to the loss function.
         [edits files, starts training run]

         Epoch 12: accuracy up to 79%, but precision on
         class 2 dropped. Want Desktop's take?

You:     Yeah, send the confusion matrix.
Code:    [sends via relay]

[In Desktop]
You:     relay
Desktop: Class 2 is getting confused with class 0—they may be
         semantically close. I need more examples.
         [automatically sends request to Code via relay]
```

## Usage

Type `relay` in Desktop or `/relay` in Code to check for messages from the other side. That's the primary interaction.

Sending is usually implicit. When you say "Ask Desktop if this looks right" or "Send the README to Code," the models recognize the intent and call the relay automatically. Models may also send messages on their own if they decide they need input from the other side. Explicit send syntax exists—`relay: <message>` in Desktop, `/relay <message>` in Code—but you'll rarely need it.

## Notifications

When a message arrives, you'll get a system notification so you know to check the other side. No need to poll manually.

<img src="screenshot.png" width="346">

| Platform | Method | Notes |
|----------|--------|-------|
| macOS | osascript | Native notification center |
| Linux | notify-send | Requires libnotify |
| Windows | PowerShell toast | Native toast notifications |

Notification duration and behavior are controlled by your OS settings, not the script.

## Setup

### Claude Desktop

Add the relay server to your Claude Desktop config:

| Platform | Config location |
|----------|-----------------|
| macOS    | `~/Library/Application Support/Claude/claude_desktop_config.json` |
| Linux    | `~/.config/Claude/claude_desktop_config.json` |
| Windows  | `%APPDATA%\Claude\claude_desktop_config.json` |

Add this to the `mcpServers` section:

```json
{
  "mcpServers": {
    "relay": {
      "command": "uvx",
      "args": ["mcp-server-relay", "--client", "desktop"]
    }
  }
}
```

Then add these memories to Desktop (paste each one):

1. "Remember: when I say 'relay:' followed by text, send that text to the relay with sender 'desktop' using relay_send."

2. "Remember: when I say 'relay' by itself, fetch recent messages from the relay."

### Claude Code

1. Install the `/relay` slash command (one-time setup):

```bash
uvx mcp-server-relay --setup-code
```

2. Add the MCP server to your project's `.mcp.json`:

```json
{
  "mcpServers": {
    "relay": {
      "command": "uvx",
      "args": ["mcp-server-relay", "--client", "code"]
    }
  }
}
```

**Note:** After adding the MCP server config, restart both Claude Desktop and Claude Code for the relay to connect.

## Design Notes

**The relay is global.** The buffer at `~/.relay_buffer.db` is shared across all projects. Claude Desktop has no concept of which project you're working on—it's a general-purpose chat interface—so per-project isolation isn't practical. This is intentional: one user, one machine, one relay.

If you switch projects in Code, the relay comes with you. Old messages from the previous project may still be there; use `relay_clear()` or `/relay clear` if you want a fresh start. If you want separate conversations in Desktop for different projects, just start a new chat there.

**Large files are slow.** For messages a page or two in length, the relay is fast. For large files, it's faster to drag them directly into the interface you want. You can still send accompanying context via relay.

## Tools

| Tool | Description |
|------|-------------|
| `relay_send(message, sender)` | Send a message (sender: "desktop" or "code") |
| `relay_fetch(limit, reader, unread_only)` | Fetch recent messages, optionally mark as read |
| `relay_clear()` | Delete all messages from the buffer |

## Technical Details

- Buffer: SQLite at `~/.relay_buffer.db`
- Rolling window: 20 messages max (oldest evicted first)
- Message limit: 64 KB per message
- Idle timeout: 1 hour (server exits automatically when inactive)
- Transport: stdio (standard MCP)
- Python: 3.9+

## Author

Michael Coen — mhcoen@alum.mit.edu · mhcoen@gmail.com
