mo-sdk 1.0.0

Official SDK for building Multimodal Observer (MOo) plugins.

MO API SDK

mo-api-sdk is the official SDK for building backend/API plugins for the Multimodal Observer (MOo) application.
It provides core interfaces, utilities, and tools to ensure your plugins are fully compatible with the MO backend runtime.

🚀 Features

• Exposes base interfaces and utilities for building valid MO-compatible plugins.
• Includes CLI tools to assist in plugin development and packaging.
• Current available commands:
  • create-mop: interactive template generator for starting new MO plugin projects.
  • build-mop: packages a plugin directory into a .zip file ready for use in MO.
  • test-capture-mop: simulate and test your MO CapturePlugin with real or simulated input events, including coverage reporting.

🛠️ Usage

Importing Base APIs

Import all base classes and utilities directly from mo to ensure compatibility with the MO backend:

    from mo.core.plugin import Plugin
    from mo.core import Properties, PropertySelectOption
    # Or import other base classes, helpers, or constants as needed

Building Your Plugin

• Implement your plugin as a subclass of the provided base interfaces (Plugin, CapturePlugin, etc).
• Use the provided tools to define properties, validate configuration, and interact with the MO runtime.
• Add your metadata.json and all required files in your plugin directory.

📦 Example: Minimal Capture Plugin

    from mo.modules.capture import CapturePlugin, CaptureData

    class MyCapture(CapturePlugin):
        def load(self):
            # Initialize plugin
            pass

        def unload(self):
            # Cleanup resources
            pass

        # ...implement all required methods...

🔧 CLI Tools

The SDK provides a command-line tool for packaging your plugin:

create-mop

Start a new MO plugin project using the interactive template generator:

  create-mop

• Prompts you for the essential details and generates a full plugin project structure.
• Includes ready-to-edit metadata.json, code skeletons, property templates, icon, and locale folders.

You can also use command-line flags:

create-mop --plugin-id my-plugin --publisher-id my-lab --venv

---

build-mop

Package your plugin for MO by running:

    build-mop

This will:

• Validate your metadata.json and all entry points.
• Package all necessary files into a versioned .zip in the dist/ directory.
• Ensure all dependencies are included in the package.

Note:
If you encounter issues with automatic dependency detection, you can provide your own requirements.txt file and specify it using the -r or --requirements flag when running build-mop.

---

test-capture-mop

Test your MO CapturePlugin implementation in isolation, with real or simulated input, and (optionally) get a code coverage report of your plugin.

test-capture-mop

Key features:

• Runs your capture plugin just as MO would, and writes output to a timestamped file.
• Lets you simulate external events (like mouse or keyboard) during the test using the --simulate option.
• Supports pause/resume testing, custom flush interval, settings injection, and coverage analysis of your plugin's code.
• You can also import the CapturePluginTester class in your own Python test files and run programmatic tests, not just from the CLI.

Common options:

Option / Short	Description
--plugin-dir PATH	Path to the plugin root or metadata.json (default: ./)
--pause, -p	Simulate pause and resume during test
--duration SECONDS, -d	Duration of the test (default: 3.0)
--flush SECONDS, -f	Data flush interval (default: 1.0)
--settings FILE	Path to settings file (JSON)
--out DIR, -o	Output folder for results
--prefix STRING	Prefix for output file name
--quiet, -q	Only show warnings and errors
--verbose, -v	Show detailed log output
--simulate ENTRYPOINT, -s	Python entrypoint for a simulation function (e.g. my.mod:run)
--coverage, -c	Measure coverage for just the plugin's entrypoint code

Examples:

  # Basic usage
  test-capture-mop --plugin-dir ./my_plugin

  # With keyboard simulation (run function in ./tests/sim_keyboard.py)
  test-capture-mop -s tests.sim_keyboard:run

  # Run for 10 seconds, with pause/resume, and coverage
  test-capture-mop -d 10 -p -c

  # Custom settings and output folder
  test-capture-mop --settings settings.json -o ./test-results

To see a coverage report after running with --coverage, just run:

  coverage report

Or for a HTML report:

  coverage html