MCP Setup

Local MCP setup for Claude, Codex, and VS Code.

DreamCleanr already ships a local MCP server. The MCP surface stays preview-first: you can scan, inspect, render reports, and preview scheduled cleanup without exposing destructive cleanup by default.

Local-first The MCP server runs on your Mac through `python3 -m dreamcleanr.mcp_server`.

Preview-first `scan`, `clean_preview`, `report_render`, and schedule tools ship now.

No fake browser magic The site demo is a demo. Real MCP scans happen locally after install.

Start here

Install once, then paste the client config you need.

1. Install DreamCleanr

Use the GitHub-first install path so the local CLI and MCP module are available on your Mac.

curl -fsSL https://raw.githubusercontent.com/jlsport18/DreamCleanr/main/scripts/install.sh | bash

2. Verify the local entrypoint

Before wiring a client, make sure DreamCleanr is installed and the MCP entrypoint resolves locally.

dreamcleanr --version
python3 -m dreamcleanr.mcp_server

If the second command starts the local server without import errors, the client configs below are the right shape.

Client configs

Copy the config that matches your client.

Claude

Reference file: `integrations/claude-mcp.json`

{
  "mcpServers": {
    "dreamcleanr": {
      "command": "python3",
      "args": ["-m", "dreamcleanr.mcp_server"]
    }
  }
}

Codex

Reference file: `integrations/codex-mcp.toml`

[mcp_servers.dreamcleanr]
command = "python3"
args = ["-m", "dreamcleanr.mcp_server"]

VS Code

Reference file: `integrations/vscode-mcp.json`

{
  "servers": {
    "dreamcleanr": {
      "command": "python3",
      "args": ["-m", "dreamcleanr.mcp_server"]
    }
  }
}

What you get over MCP today

`scan`
`clean_preview`
`report_render`
`schedule_status`
`schedule_preview`

Destructive cleanup is intentionally not exposed as a default MCP action.

Troubleshooting

Keep the setup path grounded.

If the config does not load

Check the local Python path first.

Confirm `dreamcleanr --version` works in the same shell environment.
Confirm `python3 -m dreamcleanr.mcp_server` launches without import errors.
Re-run the public installer or updater if the module is missing.

If you want more context

Use the FAQ and pricing plan next.

DreamCleanr keeps the MCP setup local and free today, while the premium shell remains a later one-time upgrade.

Read the FAQ See pricing