ezmsg-dashboard

ezmsg-dashboard is a web dashboard for inspecting and operating running ezmsg systems.

It combines:

• a Python backend that talks to GraphContext
• a React/TypeScript frontend for topology, publishers, settings, and trace visualization
• fixture-backed tests for graph layout, inspector behavior, profiling, and visual regression coverage

This package is published as ezmsg.dashboard and follows the namespace packaging style used by other ezmsg extensions.

The release artifact is intended to be installed as a normal Python package:

pip install ezmsg ezmsg-dashboard

To enable the live data viewer, install the viz extra, which adds ezmsg-tools for its AxisArray-to-plottable reduction:

pip install "ezmsg-dashboard[viz]"

Runtime installs do not require npm. The published wheel/sdist includes a prebuilt frontend bundle that the Python backend serves directly.

Documentation lives in docs/README.md.

NOTE: This software was written in heavy collaboration (vibe coded) with ChatGPT 5.4. It appears functional and has been used to evaluate a variety of ezmsg deployments, but this in no way implies fitness for any particular use, or that the code is anything more than AI slop. The human(s) who have their name associated with this package do not fully understand how the code was designed/functions and will not necessarily be helpful in GitHub issues or PRs. It should be treated as a tool that is nice when it works well and solves a problem, and as an inspirational jumping board/mockup for what ezmsg-dashboard could be with a real development push by human developers.

Features

• Interactive Build mode: discover registered components, configure settings, wire ports, save/load graphs, and run or replace a dashboard-managed processing group
• Live topology rendering with left-to-right and top-to-bottom layouts
• Scoped collection navigation with breadcrumb and in-graph open/up controls
• Settings inspection and patching
• Publisher and subscriber profiling views
• Profiling trace capture and timing visualization
• Live data visualization: GPU-rendered sweep, spectrum, and channel-map views of any publisher's messages, plus a message inspector that works on any topic
• Frontend fixture mode for deterministic graph and profiling scenarios
• Unit, Playwright, and screenshot-based regression tests

Running

End-user runtime

After installing ezmsg and ezmsg-dashboard, launch the packaged dashboard server directly:

ezmsg dashboard --graph-address 127.0.0.1:25978

This starts the Python backend and serves the packaged frontend from the same process.

If you want core ezmsg to host the graph server and dashboard together:

ezmsg serve --dashboard

For local setup, development mode, testing, fixture scenarios, and release steps, use the Development Guide.

License

MIT. See LICENSE.txt.