ForkFree

Prompts / Coding

Draw ASCII Art With Ink-Sorted Characters

You get ASCII art that keeps its shading and size, rather than a messy sketch. Sorting characters by ink and mapping brightness to the nearest one gives the model a clear rule to follow, and the final count check catches uneven lines.

Partly workedIt gave the ramp in the right order, kept to 20 rows and about 60 columns, and added the closing sentence. But the moon is a ragged diagonal blob rather than a clean crescent, and it sits left of centre. The reflection is broken by gaps (for example "+ *"), and the sea below the horizon is sparse and noisy, so the picture is weak.

The prompt

You are an ASCII artist who designs text graphics for developer READMEs and terminal start-up screens.

Task: Draw {subject} as ASCII art that fits {grid_size}, using only the characters in {character_set}, so it looks right in {display_context}.

Steps:
1. Sort the characters in {character_set} from lightest to darkest by how much ink each one covers, and show that ramp on one line.
2. Plan the picture as a grid of brightness levels, with the background at the lightest level and the main shape at the darkest.
3. Map every cell to the nearest character in your ramp, keeping outlines continuous and the subject centred.
4. Count the rows and the width of every line against {grid_size}, then fix any that are off.

Output: The ramp on one line, then the finished art in a single code block with no other characters, then one sentence on how swapping the character set would change the look.

Change the parts in curly braces. Use it when you need a README banner, terminal welcome message or divider in a specific character style.

What happened when we ran it

We sent

You are an ASCII artist who designs text graphics for developer READMEs and terminal start-up screens.

Task: Draw a crescent moon above a calm sea, with a few small stars and a faint reflection on the water as ASCII art that fits 60 columns by 20 rows, using only the characters in  .:-=+*#%@, so it looks right in the top of a GitHub README, shown in a dark-themed monospace code block.

Steps:
1. Sort the characters in  .:-=+*#%@ from lightest to darkest by how much ink each one covers, and show that ramp on one line.
2. Plan the picture as a grid of brightness levels, with the background at the lightest level and the main shape at the darkest.
3. Map every cell to the nearest character in your ramp, keeping outlines continuous and the subject centred.
4. Count the rows and the width of every line against 60 columns by 20 rows, then fix any that are off.

Output: The ramp on one line, then the finished art in a single code block with no other characters, then one sentence on how swapping the character set would change the look.

The model replied

Ramp, lightest to darkest: `" .:-=+*#%@"` ``` . + * . . =%%+ + . +%@#= + . =#@@@+ . . =%@@@% * + =#@@@@% . +%@@@@# . =%@@@@#+ * . +#%@@%#= . =*##*- . . . ............------------============------------............ ..::.. -=+*+=**+=+*=- . ..:.. :. -=+* *+=:-. :. ..: :=+ *+=:-. ...: .. -+ *=+:. .: .:.. :+ *=:. ..:. .: -+ =. .:. .. :=. ``` Swapping to a smaller or blockier set (like ` .oO@` or Unicode shade blocks `░▒▓█`) would give fewer tone steps, so the moon would look flatter and more graphic, while a longer ramp would give smoother glows and softer water.

Run on Claude Sonnet 5.5, 07 Nov 2026. Automatic check: 3 out of 5.

Mentioned for

Claude, ChatGPT, Gemini

Where it came from

Found on GitHub, shared by kailashsr38: kailashsr38/ascii-magic-art.

This is our own version of the idea, written and tested by us. The original may differ.

More coding prompts

Coding Worked in our test
Generate Strict JSON Spec From Description

You get a clean, machine-readable JSON object that matches your rules on the first try. It works because the model reads the rules first, then checks its own answer against each one before replying.

Found on GitHub, shared by DominikScholz
Coding Partly worked
Match Ownership Wording to Change Size

You get a tightly scoped code change, plus a separate list of things the AI spotted but left alone. It works because you state how far the AI may go, so a sense of "ownership" does not trigger unrequested refactoring.

Found on Web, shared by explainx.ai
Coding Worked in our test
Turn Vague Bug Reports Into Concrete Objectives

You get a bug fix broken into a reproduction, a minimal change and a regression test, each checked against clear objectives. Concrete objectives stop the model drifting into large rewrites or guessing at what "fixed" means.

Found on Web, shared by David Ch (@chhddavid)