oneshotlm
Compare all models on this prompt →

Reaction-diffusion · seed-2.0-code

Run a Gray-Scott reaction-diffusion simulation on a canvas that grows organic Turing patterns in real time, with the mouse seeding new growth.

⛨ sandboxed Open in new tab ↗
transcript · 12 messages
user

Run a Gray-Scott reaction-diffusion simulation on a canvas that grows organic Turing patterns in real time, with the mouse seeding new growth.

assistant
thinking
Let me create a Gray-Scott reaction-diffusion simulation as requested. This is a well-known pattern formation system that models chemical reactions. I'll implement it on an HTML5 canvas with real-time simulation, mouse interaction, and responsive design.

Key features needed:
1. Canvas-based simulation using 2D context
2. Two floating-point grids (current and next state) for chemicals A and B
3. Laplacian operator for diffusion
4. Reaction kinetics (feed and kill rates)
5. Mouse interaction to seed B concentrations
6. Color mapping for visualization
7. Performance optimizations

Let me write this as a single self-contained HTML file.
write /tmp/ws/index.html (17998 bytes)
toolResult

Successfully wrote 17998 bytes to /tmp/ws/index.html

assistant
thinking
The file has been created successfully. Let me verify it exists and is well-formed by checking its contents briefly.
bash {"command":"ls -la /tmp/ws/index.html"}
toolResult

-rw-r--r--. 1 node node 18000 Aug 13 01:32 /tmp/ws/index.html

assistant
thinking
The file is present. Let me now start a simple web server to serve it, so the user can view the simulation.
bash {"command":"cd /tmp/ws && python3 -m http.server 52197","timeout":3600}
toolResult

/bin/bash: line 1: python3: command not found Command exited with code 127

assistant
thinking
Python3 is not available. Let me try with node to create a simple server, or check what's available.
bash {"command":"which node npx python2"}
toolResult

/usr/local/bin/node /usr/local/bin/npx Command exited with code 1

assistant
thinking
Node is available. Let me create a simple HTTP server using Node.js to serve the index.html file.
write /tmp/ws/server.js (842 bytes)
toolResult

Successfully wrote 842 bytes to /tmp/ws/server.js

assistant
thinking
Now let me start the Node.js server to serve the simulation.
bash {"command":"cd /tmp/ws && node server.js","timeout":3600}

TAKE IT TO CODE

A single self-contained HTML file — copy it and open in any browser, or reuse the prompt with your own pi-agent run.

RUN

modelseed-2.0-code
vendorByteDance · Seed
harnesspi-agent
generated2026-08-13

EVALUATION

4/5 working

Gray-Scott simulation grows organic maze-like Turing patterns and responds well to clicks and drags seeding growth.

Vision grade · anthropic/claude-sonnet-5

MEASUREMENTS

Raw readings from the capture, not a verdict — the grade above is the only judgement. Percentages are how much of the screen changed; the measure under-reports sparse motion on a dark background.

idle motion0.2%
on click17.3%
on drag11.7%
on wheel4.6%
on enter + space3.1%
on arrow keys2.0%
on w/a/s/d1.7%
frame spread21.5 / 255
console errors0
js errors none

METRICS

runtime1200.0s
tokens in39k
tokens out6.0k
cached26k
cost / run$0.0241