lux

A visual output surface for AI agents.

Lux gives agents and apps a shared visual surface. The intended architecture is a hub/display split: clients send UI descriptions to luxd, the Hub owns authoritative element state and behavior, and the Display renders a replica of the current scene while forwarding user interactions back to the Hub.

The design draws on X11's client/server split and Smalltalk-style live introspection. MCP is one gateway into Lux, not the whole architecture. If you want the short version of the rewrite target, start with docs/architecture/target/target.md. If you need help navigating the docs, use docs/README.md.

Platforms: macOS, Linux

Stage: alpha --- protocol is stable, published on PyPI as punt-lux

A Claude Code plugin displaying a project issue board --- the agent fetches live data from DoltDB via bd list --json, builds a filterable table with detail panel, and renders it in a single tool call. Filters and row selection run at 60fps with zero MCP round-trips.

Beads issue board with filterable table and detail panel

The same list/detail pattern generalizes to any tabular data. Search, combo filters, pagination, and a detail panel --- all driven by a single show_table() call.

Filterable data explorer with detail panel

Dashboards compose metric cards, charts, and tables. show_dashboard() builds the layout from structured data --- no manual element positioning needed.

Dashboard with metrics, charts, and data table

Quick Start

curl -fsSL https://raw.githubusercontent.com/punt-labs/lux/71014e1/install.sh | sh

Restart Claude Code twice. The Lux display window opens automatically when agents send visual output.

Manual install (if you already have uv)

uv tool install 'punt-lux[display]'

Then install the plugin via the marketplace:

claude plugin marketplace add punt-labs/claude-plugins
claude plugin install lux@punt-labs

Lightweight install (library use only)

If you only need DisplayClient to send scenes from Python (no display server):

uv add punt-lux

This pulls ~2 MB of lightweight deps. The 66 MB display stack (imgui-bundle, numpy, Pillow, PyOpenGL) is only needed for lux display and is available via punt-lux[display].

Verify before running

curl -fsSL https://raw.githubusercontent.com/punt-labs/lux/71014e1/install.sh -o install.sh
shasum -a 256 install.sh
cat install.sh
sh install.sh

Run a demo

lux display &
uv run python demos/dashboard.py

Demos are in demos/ --- each connects as a client and drives the display:

Demo	What it shows
`interactive.py`	Sliders, checkboxes, combos, text inputs, color pickers
`containers.py`	Windows, tab bars, collapsing headers, groups
`dashboard.py`	Multi-window layout with draw canvases and live controls
`data_viz.py`	Tables, plots, progress bars, spinners, markdown
`menu_bar.py`	Custom menus, event handling, periodic refresh

Features

lux

A visual output surface for AI agents.

Platforms: macOS, Linux

Stage: alpha --- protocol is stable, published on PyPI as punt-lux

Beads issue board with filterable table and detail panel

The same list/detail pattern generalizes to any tabular data. Search, combo filters, pagination, and a detail panel --- all driven by a single show_table() call.

Filterable data explorer with detail panel

Dashboards compose metric cards, charts, and tables. show_dashboard() builds the layout from structured data --- no manual element positioning needed.

Dashboard with metrics, charts, and data table

Quick Start

curl -fsSL https://raw.githubusercontent.com/punt-labs/lux/71014e1/install.sh | sh

Restart Claude Code twice. The Lux display window opens automatically when agents send visual output.

Manual install (if you already have uv)

uv tool install 'punt-lux[display]'

Then install the plugin via the marketplace:

claude plugin marketplace add punt-labs/claude-plugins
claude plugin install lux@punt-labs

Lightweight install (library use only)

If you only need DisplayClient to send scenes from Python (no display server):

uv add punt-lux

This pulls ~2 MB of lightweight deps. The 66 MB display stack (imgui-bundle, numpy, Pillow, PyOpenGL) is only needed for lux display and is available via punt-lux[display].

Verify before running

curl -fsSL https://raw.githubusercontent.com/punt-labs/lux/71014e1/install.sh -o install.sh
shasum -a 256 install.sh
cat install.sh
sh install.sh

Run a demo

lux display &
uv run python demos/dashboard.py

Demos are in demos/ --- each connects as a client and drives the display:

Demo	What it shows
`interactive.py`	Sliders, checkboxes, combos, text inputs, color pickers
`containers.py`	Windows, tab bars, collapsing headers, groups
`dashboard.py`	Multi-window layout with draw canvases and live controls
`data_viz.py`	Tables, plots, progress bars, spinners, markdown
`menu_bar.py`	Custom menus, event handling, periodic refresh

Help us improve

Find plugins for your project

Help us improve

lux-dev

Popularity

Health & Quality

Confidence

What's Inside

Help us improve

README

lux

Quick Start

Features

Similar Plugins

vizro-e2e-flow

data-visualization-creator

apple-bento-grid

fluxwing-skills

html-skills

tonone-lens

More by punt-labs

prfaq-dev

z-spec

beadle-dev

ethos-dev

biff-dev

lux

Quick Start

Features