Install and verify MirrorNeuron

Prerequisites

Install and start Docker. Local models additionally need Docker Model Runner and compatible hardware; blueprints that use an external provider may have different requirements. For source development, use Python 3.11 or newer and the deployment repository's local setup.

Install from a deployment checkout

From your checkout of mn-deploy, inspect the installer options and run it:

./install.sh --help
./install.sh

The installer uses default selections unless --interactive is supplied. Select a matching released component set with --version <release-tag>. Contributors using sibling runtime source repositories can use ./install.sh --mode local.

Installation writes local runtime configuration and installs components. Review installer options before running it on an existing installation.

Start and verify

From any directory after installation:

mn --version
mn --help
mn runtime start
mn runtime status
mn node list

Startup should report Runtime node ready. Resolve failed required components before launching a blueprint. If a check fails, run mn runtime doctor and follow Troubleshooting.

Startup displays a federation join token. Keep it private; rotate an exposed token with mn node refresh-token.

Find your services and data

Item	Default
REST API	`http://localhost:54001/api/v1`
Core gRPC	`localhost:55051`
Web UI	`http://localhost:55173`
Local runtime state	`~/.mn`
Blueprint run records	`$MN_HOME/runs/<run-id>`
Downloaded run outputs	`$MN_HOME/outputs/<run-id>`

Use mn runtime status to find the actual configured endpoints. Configuration explains overrides. Do not share generated environment files or credentials.

Stop services

When other runs no longer need the local runtime:

mn runtime stop

Keep run records and configuration needed for recovery. Removing ~/.mn can discard runtime state; stopping services does not require deleting it.

Continue with Run your first blueprint.