Skip to the content
2 layers160 × 70 mmPanel, the printed circuit board in 3D

Engiworld: What Can Frontier Agents Deliver in Professional Engineering Environments?

Size
160 × 70 mm
Copper layers
2
Thickness
1.6 mm
Parts
30
Nets
3
Pads
48
Tracks, vias
0, 0
Routed
Not yet
KiCad
KiCad 9.0 (file 20240108)
Created
12 Apr 2026
Updated
29 Oct 2026

Overview

Panel is an open-source printed circuit board by Hongcheng-Gao, from the Hongcheng-Gao/EngiWorld repository on GitHub, published under the MIT licence. It is a 2-layer board of 160 × 70 mm with 30 parts on 3 nets.

README

README.md

English | 简体中文

EngiWorld

What Can Frontier Agents Deliver in Professional Engineering Environments?

🌐 Website · 📄 Paper · 🧩 Tasks & Verifiers · 🤗 Task Dataset · 🤗 Environment Images

MIT License

EngiWorld task examples and engineering software coverage

Six task types for engineering software use and artifact-based evaluation. Click to enlarge.

EngiWorld evaluates whether agents can operate professional engineering software and deliver the requested engineering artifacts. This repository provides all 1,301 tasks, their evaluators, and the GUI/CLI evaluation runtime. The paper uses a 306-task main-experiment subset, listed in task/splits/main-306.txt.

The full set of 1,301 tasks and their verifiers is also available in the Hugging Face task dataset, with a dataset viewer for browsing task instructions and metadata.

The benchmark covers CAD, CAE, CAM, BIM, EDA, and DCC. Its six task types are Single-Software Execution (931), Software Selection (140), Vision-Guided Modeling (120), Design Optimization (40), Cross-Software Coordination (60), and Open-Environment Engineering (10).

Local evaluation launches Windows or Ubuntu virtual machines through Docker on Linux, calls the model, and saves evaluation results.

☁️ Run on Alibaba Cloud or Volcengine

Use the cloud evaluation guide for single-task or Redis-scheduled batch evaluation on either cloud. Explicit model presets, API interruption handling and batch state management are shared. Images require compatible engineering applications and desktop services.

💾 Installation

Prepare an x86_64 Linux server with Python 3.10, Docker Engine, and zstd. With /dev/kvm, the VM uses hardware acceleration. Without it, the Docker runner attempts software emulation, which starts and runs more slowly. Each environment defaults to 4 CPUs and 16 GB RAM.

Get the code and enter the runtime directory:

bash
git clone --branch main --single-branch --depth 1 https://github.com/Hongcheng-Gao/Engiworld.git &&
cd Engiworld/runtime

Install the runtime in a Python venv environment:

bash
python3.10 -m venv .venv
source .venv/bin/activate
python -m pip install -e ".[worker-api,download]"

Run the following commands from runtime/.

💾 Download complete data before evaluation

The complete dataset provides all 1,301 task definitions, drawings, evaluators, and required ground-truth files. GitHub includes every drawing and evaluator plus smaller task resources. Download the complete data before starting evaluation.

SourceComplete task dataEnvironment images
Hugging FaceEngiWorld-TasksEngiWorld-Images
ModelScopeEngiWorld-TasksEngiWorld-Images

Use --source modelscope in the commands below to download from ModelScope.

Hugging Face:

bash
python -m engiworld.prepare_tasks --source hf --task-root ../data/task
python -m engiworld.prepare_images --source hf --image LibreCAD2.2.0.2 --image-root ../environment-images

ModelScope:

bash
python -m engiworld.prepare_tasks --source modelscope --task-root ../data/task
python -m engiworld.prepare_images --source modelscope --image LibreCAD2.2.0.2 --image-root ../environment-images
bash
zstd -d ../environment-images/images/LibreCAD2.2.0.2.qcow2.zst
docker pull happysixd/osworld-docker
python -m engiworld.prepare_tasks --task-root ../data/task --check

Downloads verify checksums and preserve modified local files. ModelScope image downloads use 8 parallel range streams by default; set MODELSCOPE_DOWNLOAD_PARALLEL_WORKERS to adjust this. Omitting --image downloads all 15 public archives (about 98 GB compressed). Choose the image matching the task snapshot from manifest.json. Snapshots outside the public catalog require a matching image prepared by the user. Decompress the selected archive and prepare the Docker runner before evaluation. Task preparation uses the published dataset directly.

🚀 Start from local data

The following commands read already downloaded task data and the local image.

1. Start a task

bash
python -m engiworld.local_eval \
  --task-root ../data/task \
  --task single-software-execution/gui/librecad/task-08 \
  --image ../environment-images/images/LibreCAD2.2.0.2.qcow2 \
  --smoke-test

This starts the VM, loads the task inputs, and saves a smoke.png screenshot in the results directory without calling a model.

Servers without KVM: Check whether the server provides /dev/kvm:

bash
test -e /dev/kvm && echo "KVM available" || echo "Software emulation"

If the output says Software emulation, run the same --smoke-test command above. The runtime selects software emulation automatically; no arguments need to change. Startup may take several minutes. When the terminal shows "status": "environment_ready" and smoke.png appears in the results directory, open the screenshot to confirm that LibreCAD and the input drawing are ready. Then follow step 2 to run and score a model.

2. Run a model and score the task

Configure an OpenAI-compatible Chat Completions endpoint and run the task:

bash
export OPENAI_BASE_URL="https://your-api-endpoint/v1"
export OPENAI_API_KEY="your-api-key"

python -m engiworld.local_eval \
  --task-root ../data/task \
  --task single-software-execution/gui/librecad/task-08 \
  --image ../environment-images/images/LibreCAD2.2.0.2.qcow2 \
  --model your-model-name \
  --result-dir results

When the run finishes, summary.json summarizes the run, result.txt holds the score, and traj.jsonl records model actions. For another task, change --task and use its matching local image. Complete-data preparation already supplies its resources. See the runtime guide for more options.

Repository layout

text
task/<category>/<gui|cli>/<software or software1--software2>/<task-number>/
runtime/

Task categories are single-software-execution, cross-software-coordination, software-selection, design-optimization, vision-guided-modeling, and open-environment-engineering. Open-Environment Engineering tasks use CLI in a blank environment and live under task/open-environment-engineering/cli/agent-selected/; the agent chooses and installs its tools.

Task JSON IDs follow <category>--<gui|cli>--<software>--<task-number>--<os>, for example single-software-execution--gui--librecad--task-08--ubuntu. Use either this ID or its task directory path with --task. See the task guide.

Directory categories and JSON IDs follow the updated paper taxonomy as of 2026-10-09. task/aliases.json preserves historical JSON IDs and directory names; the runtime continues to accept them for task selection. See the full ID migration map and task taxonomy.

📊 Evaluation results

The main experiment evaluates seven models on the same 306 tasks: 158 CLI and 148 GUI, including all 10 Open-Environment Engineering tasks. EngiScore is the mean task score on a 0–100 scale. Execution metrics average all 306 tasks, including unsuccessful runs; output tokens are reported per turn, in thousands.

ModelEngiScoreCLIGUITurns/taskTokens (K/turn)Cost ($/task)
Claude Opus 544.748.640.569.52.418.78
GPT-5.6 Sol38.647.928.654.21.09.57
Qwen3.8 Max26.335.316.6107.32.76.44
Gemini 3.7 Flash26.137.214.2110.00.72.08
DeepSeek V4.1 Flash25.948.32.0115.39.80.80
Kimi K324.331.316.792.10.95.85
Qwen3.8 Flash16.731.51.0119.61.30.44

📚 Citation

If you use EngiWorld in your research, please cite:

bibtex
@misc{engiworld2026,
  title = {EngiWorld: What Can Frontier Agents Deliver in Professional Engineering Environments?},
  author = {Hongcheng Gao and Hailong Qu and Yu Lei and Henghui Sun and Haoyang Li and Yipeng Wei and Naihao Xue and Xiaohan Yu and Zhuo Tao and Yihe Zang and Yajiao Wang and Jingyi Tang and Yi Li and Jingjing Zhou and Jie Luo and Bohan Zeng and Chengyu Shen and Hao Jiang and Chong Chen and Bowen Qu and Olive Huang and Zeqiang Wang},
  year = {2026},
  eprint = {2609.37686},
  archivePrefix = {arXiv},
  primaryClass = {cs.AI},
  url = {https://arxiv.org/abs/2609.37686}
}

📄 License

EngiWorld's original code is released under the MIT License. See Third-party notices for component licenses.

Parts

30
QtyValueFootprintReferences
12FiducialFiducial:Fiducial_1mm_Mask2mmFID1, FID2, FID3, FID4, FID5, FID6, FID7, FID8, FID9, FID10, FID11, FID12
6100nFCapacitor_SMD:C_0402A1_C1, A2_C1, A3_C1, A4_C1, A5_C1, A6_C1
610kResistor_SMD:R_0402A1_R1, A2_R1, A3_R1, A4_R1, A5_R1, A6_R1
6LED_GLED_SMD:LED_0402A1_D1, A2_D1, A3_D1, A4_D1, A5_D1, A6_D1

More from Hongcheng-Gao/EngiWorld

12
  • 2 layers146 × 110 mmPanel, the printed circuit board in 3D

    Panel

    Hongcheng-Gao/EngiWorld

    Engiworld: What Can Frontier Agents Deliver in Professional Engineering Environments?

    29 parts52MIT
  • 2 layers109 × 78 mmPanel, the printed circuit board in 3D

    Panel

    Hongcheng-Gao/EngiWorld

    Engiworld: What Can Frontier Agents Deliver in Professional Engineering Environments?

    22 parts52MIT
  • 2 layers165 × 52 mmStage3 terminal cover long board, the printed circuit board in 3D

    Stage3 terminal cover long board

    Hongcheng-Gao/EngiWorld

    Engiworld: What Can Frontier Agents Deliver in Professional Engineering Environments?

    9 parts52MIT
  • 2 layers88 × 70 mmSeed board, the printed circuit board in 3D

    Seed board

    Hongcheng-Gao/EngiWorld

    Engiworld: What Can Frontier Agents Deliver in Professional Engineering Environments?

    8 parts52MIT
  • 2 layers80 × 60 mmGround truth, the printed circuit board in 3D

    Ground truth

    Hongcheng-Gao/EngiWorld

    Engiworld: What Can Frontier Agents Deliver in Professional Engineering Environments?

    8 parts52MIT
  • 2 layers65 × 35 mmGround truth, the printed circuit board in 3D

    Ground truth

    Hongcheng-Gao/EngiWorld

    Engiworld: What Can Frontier Agents Deliver in Professional Engineering Environments?

    8 parts52MIT
  • 2 layers140 × 110 mmInit file, the printed circuit board in 3D

    Init file

    Hongcheng-Gao/EngiWorld

    Engiworld: What Can Frontier Agents Deliver in Professional Engineering Environments?

    8 parts52MIT
  • 2 layers80 × 80 mmInit file, the printed circuit board in 3D

    Init file

    Hongcheng-Gao/EngiWorld

    Engiworld: What Can Frontier Agents Deliver in Professional Engineering Environments?

    8 parts52MIT
  • 2 layers80 × 52 mmSeed board, the printed circuit board in 3D

    Seed board

    Hongcheng-Gao/EngiWorld

    Engiworld: What Can Frontier Agents Deliver in Professional Engineering Environments?

    7 parts52MIT
  • 2 layers96 × 58 mmOptimized board, the printed circuit board in 3D

    Optimized board

    Hongcheng-Gao/EngiWorld

    Engiworld: What Can Frontier Agents Deliver in Professional Engineering Environments?

    7 parts52MIT
  • 2 layers110 × 66 mmOptimized board, the printed circuit board in 3D

    Optimized board

    Hongcheng-Gao/EngiWorld

    Engiworld: What Can Frontier Agents Deliver in Professional Engineering Environments?

    7 parts52MIT
  • 2 layers76 × 76 mmOptimized board, the printed circuit board in 3D

    Optimized board

    Hongcheng-Gao/EngiWorld

    Engiworld: What Can Frontier Agents Deliver in Professional Engineering Environments?

    7 parts52MIT

Similar boards

  • 2 layers52.2 × 52.2 mmBase promicro, the printed circuit board in 3D

    Base promicro

    google/mozc-devices

    Circuit diagrams and firmware source code for Gboard DIY keyboards

    30 parts3kApache-2.0
  • 2 layers40.6 × 15.2 mmVr mmm, the printed circuit board in 3D

    Vr mmm

    rusefi/rusefi

    rusefi - GPL internal combustion engine control unit

    30 parts1.2k
  • 2 layers270 × 11 mmDaylight, the printed circuit board in 3D

    Daylight

    VoronDesign/Voron-Hardware

    Cool Hardware designed by the Voron Design group

    30 parts1.1kGPL-3.0
  • 2 layers65 × 30 mmRgb to hdmi, the printed circuit board in 3D

    Rgb to hdmi

    hoglet67/RGBtoHDMI

    Bare-metal Raspberry Pi project that provides pixel-perfect sampling of Retro Computer RGB/YUV video and conversion to HDMI

    30 parts1kGPL-3.0
  • 2 layers40.5 × 58 mmInitial v, the printed circuit board in 3D

    Initial v

    tenderlove/initial-v

    It's a BMW shifter converted to a Bluetooth Keyboard that you use with Vim

    30 parts1kApache-2.0
  • 2 layers34.3 × 41.8 mmMotionCubeViewAllForces, the printed circuit board in 3D

    MotionCubeViewAllForces

    hackclub/OnBoard

    💡 Join 1,000 teenagers and make your first PCB with a $100 grant!

    30 parts915
  • 2 layers57.4 × 62 mmPicoWS2812, the printed circuit board in 3D

    PicoWS2812

    ErichStyger/mcuoneclipse

    McuOnEclipse projects and examples

    30 parts765
  • 2 layers52.1 × 46.8 mmFIRE TEST PANEL, the printed circuit board in 3D

    FIRE TEST PANEL

    jrsteensen/OpenHornet

    OpenHornet 1:1 F/A-18C Simulator Repository

    30 parts431

Questions

How big is the Panel PCB?

It measures 160 × 70 mm and has 2 copper layers, 1.6 mm thick.

What parts does the Panel use?

30 parts. The full list of its parts is on this page, with links to each part's symbol and footprint where PCBKit's library has them.

Where can I download the Panel KiCad files?

From the Hongcheng-Gao/EngiWorld repository on GitHub, at the commit this page was made from: the board is task/single-software-execution/cli/kicad/task-19/ground_truth/panel.kicad_pcb. PCBKit opens the board in the browser, with nothing to install.

Can I use the Panel design in my own project?

Its repository publishes it under the MIT licence: you may reuse it on that licence's terms (keep the author's notice, and for a copyleft licence share your changes alike).