explore-variations

MCP prompt for exploring parameter and seed variations of an existing sketch

explore-variations

Structured prompt template that guides an AI agent through exploring variations of an existing sketch by adjusting parameters and seeds.

Arguments

Argument	Type	Required	Description
`sketchId`	`string`	yes	ID of the sketch to explore variations of
`strategy`	`string`	no	Variation strategy: `random`, `systematic`, `gradient`
`count`	`number`	no	Number of variations to generate

Strategies

Strategy	Description
`random`	Randomize all parameters and seed for each variation
`systematic`	Vary one parameter at a time across its range
`gradient`	Gradually shift parameters in a direction

Behavior

The prompt guides the agent through:

Read the source sketch — get current parameters, seed, and algorithm
Fork variations — create copies with modified parameters/seeds
Arrange on canvas — lay out variations in a grid for visual comparison
Capture thumbnails — render all variations for side-by-side review

Example usage

Use the explore-variations prompt with:
  sketchId: "spiral-particles"
  strategy: "systematic"
  count: 6

Output

The prompt guides the agent to use MCP tools in sequence:

get_selection or open_sketch — load the source sketch
fork_sketch — create N copies with parameter/seed changes
set_parameters / set_seed — adjust each variation
auto_arrange — lay out variations in a grid
capture_batch — render thumbnails for comparison

fork_sketch — the tool used to create variations
randomize_parameters — random parameter generation
auto_arrange — automatic layout

create-generative-art

MCP prompt for creating a new piece of generative art with structured guidance

apply-design-theory

MCP prompt for applying design theory concepts to improve a sketch

On this page

explore-variations Arguments Strategies Behavior Example usage Output Related