SkillAgentSearch skills...

bricky

A compiler for LEGO® models, built for AI agents.

Install / Use

claude mcp add maxencefaldor -- npx -y github:maxencefaldor/bricky

If the server publishes to npm under a different name, use that package instead — check the repo README.

About this skill
🔌

MCP Server

Model Context Protocol server

Quality Score

71/100

Supported Platforms

Claude Code
Claude Desktop

Tags

Our assessment of bricky

bricky scores 71/100 on our quality scale, 3827th of 4,656 Development & Engineering skills we index.

Its MCP Server is 3.8 KB long, split into 5 sections with 4 code examples: a solid amount of guidance for an agent.

It has 10 GitHub stars, so there is little community track record yet; judge it on its content.

Substance
26/30
Structure
18/20
Description
8/15
Adoption
4/20
Freshness
15/15

Maintenance, license and trust

  • The repository was last updated yesterday, so bricky is actively maintained.
  • It is released under the Apache-2.0 license, a permissive license that allows use, modification and commercial use with attribution.
  • Its trust signals score 92/100, with 1 caution from licensing, adoption, age or documentation. These come from repository metadata, not a code audit — read the skill file before letting an agent act on it.

bricky compared with similar skills

All 4 of these similar skills score higher than bricky; compare them before choosing.

SkillScoreStarsUpdatedFormat
bricky (this skill)by maxencefaldor71101d agoMCP Server
Agent-Reachby Panniantong100107.2k6d agoCLAUDE.md
taste-skillby Leonxlnx100106.9ktodayCLAUDE.md
headroomby headroomlabs-ai10078.5k1d agoCLAUDE.md
ai-job-searchby ai-job-search-dev10068.4ktodayCLAUDE.md

Frequently asked questions

How do I install bricky?
Run claude mcp add maxencefaldor -- npx -y github:maxencefaldor/bricky. The install tabs above show the steps for each supported agent.
Which AI agents does bricky work with?
It is written for Claude Code and Claude Desktop, as a MCP Server file. Other agents that read the same format can often use it too.
Is bricky safe to use?
It is Apache-2.0-licensed and scores 92/100 on trust signals. Skills are instructions an agent will follow, so read the file before installing it and do not approve commands you do not understand.
Is bricky still maintained?
The repository was last updated yesterday, so bricky is actively maintained.

bricky

A compiler for LEGO® models, built for AI agents.

The Japanese garden example: a pagoda, a torii and cherry trees in bloom around a koi pond

An AI agent writes a LEGO® model as a Python build script, and bricky builds it the way a compiler builds code. Each build places every part in exact geometry, reports each mistake with the script line behind it, and renders four views with rulers.

The documentation explains every concept and lists every command.

🚀 Quick start

The quick start needs uv, which installs Python if needed. A uv project holds your build scripts, and uv add installs bricky into it from PyPI. The project pins Python 3.14, the version bricky recommends: on 3.15, installing compiles the solver HiGHS, which needs a C++ compiler.

uv init --bare --pin-python --python 3.14 my-models  # a project for your build scripts
cd my-models
uv add bricky                                        # bricky and every dependency, from PyPI
uv run playwright install chromium                   # the browser bricky renders in, once per machine

Download an example and build it, which checks the model and renders it beside the script:

curl -O https://raw.githubusercontent.com/maxencefaldor/bricky/main/examples/japanese_garden.py
uv run bricky build japanese_garden.py

With pip on Python 3.12 to 3.14, run pip install bricky and then playwright install chromium. On Linux, playwright install --with-deps chromium also installs the browser's system libraries. To install bricky as a tool, run uv tool install --python 3.14 bricky --with-executables-from playwright, then playwright install chromium. The command is then bricky, without uv run. A clone of the repository runs every example and the benchmark: Get started shows how.

🤖 Use it with an agent

An agent reaches bricky through one of three front ends:

  • The command line, for any harness that runs shell commands: the agent runs uv run bricky build and reads the report and the images.
  • The MCP server, for any harness that speaks the Model Context Protocol: claude mcp add bricky -- uvx --python 3.14 bricky-mcp adds it to Claude Code.
  • The skill, which teaches the agent the build loop: Connect an agent shows how to install it.

🧱 How it works

A build script places parts by their BrickLink id, in studs and plates. The four lines below come from examples/broken.py, which makes three mistakes: an overlap, a brick half a stud off and a floating plate.

builder.place("3001", at=(0, 0, 0), color="Red", id="base")  # Brick 2 x 4
builder.place("3001", at=(1, 0, 0), color="Blue", id="overlap")  # shares space
builder.place("3003", at=(0.5, 0, 3), color="White", id="offset")  # half a stud off
builder.place("3022", at=(4, 2, 9), color="Tan", id="floating")  # touches nothing

The report reads like compiler output, and every element it names points back at the script:

error[overlap]: base and overlap occupy the same space
  --> x 1..2  y 0..4  z 0..3
   = base: Brick 2 x 4 (3001), Red, examples/broken.py:15 in build()
   = overlap: Brick 2 x 4 (3001), Blue, examples/broken.py:16 in build()

The image marks the same elements in red, so the agent sees where they are:

The broken example: every element a diagnostic names is tinted red and labeled

ℹ️ Disclaimer

LEGO® is a trademark of the LEGO Group, which does not sponsor, authorize or endorse this project.

Related Skills

View on GitHub
GitHub Stars10
CategoryDevelopment
Updated1d ago
Forks0

Languages

Python

Trust signals

92/100

From repository metadata: license, adoption, age and documentation. Not a code audit — see the Safety scan above for what the skill file itself contains.

1 low1 info