Skip to content

Adapt Generative Art¶

Adapt Generative Art — mvp.adapt_generative_art

Cluster: Creative & Media | Type: component | MCP Tools: 50

Overview¶

Generative art block that produces procedural imagery across multiple techniques — L-systems, cellular automata, fractals (Mandelbrot/Julia), context-free grammars, evolutionary art, Perlin noise, strange attractors, reaction-diffusion, particle systems, and CPPN — across image, SVG, OBJ, audio, and animated GIF modalities. Supports custom colour palettes, post-processing filter chains, layer composition, and optional headless PNG rendering via Puppeteer. Can also accept a goal context string to drive LLM-powered brief generation through JS/GLSL backends.

Launch readiness caveat

The core component test suite covers local generation, rendering, backend, persistence, and MCP paths. For first-user launch use, still run a clean-machine MCP smoke test before presenting this as frictionless: install the MCP server, run a simple generation such as generate_noise, save the result with artwork_save, and verify it can be reopened or previewed with artwork_preview. Optional paths such as Manim, Puppeteer/headless rendering, and LLM-assisted backend generation add useful capability but also add install and environment friction for non-technical users.

When to use:

Generating procedural visuals or animations from mathematical or algorithmic descriptions
Producing L-system plant or fractal imagery for synthetic training datasets or creative outputs
Composing multiple generative layers with post-processing (blur, contrast, tone-mapping)
Creating self-contained HTML/JS bundles for interactive generative sketches

Example:

from mvp.adapt_generative_art import AdaptGenerativeArtBlock, GenerativeInput

block = AdaptGenerativeArtBlock(name="gen_art")
result = block.infer(GenerativeInput(
    op="generate_lsystem",
    technique="lsystem",
    axiom="F",
    rule_string="F=F+F-F-F+F",
    iterations=4,
    width=512,
    height=512,
    palette="viridis",
))
# result.value.data_base64 → base64 PNG; result.value.svg_text → SVG markup

Works well with: adapt_image, adapt_blender, adapt_comfyui

Public API¶

`GenerativeArtPatternRuntime`¶

Stateless executable mechanism for adapt_generative_art's applied patterns.

Methods:

`review_recommendation(params: ArtParameters, strategy: str) -> tuple[str, ...]`¶

Return advisory provenance / fail-safe notes for one translation.

`AdaptGenerativeArtBlock(AIBlock[GenerativeInput, GenerativeOutput, None])`¶

Generative art block — routes by technique, converts to requested modality.

Field	Type	Default
`name`	`str`	'adapt_generative_art'
`resource_bounds`	`ResourceBounds \\| None`	None
`usage`	`ResourceUsage`	field(default_factory=ResourceUsage)

Methods:

`infer(data: GenerativeInput) -> Result[GenerativeOutput]`¶

`GenerativeInput(BaseModel)`¶

Input to AdaptGenerativeArtBlock.

Field	Type	Default
`op`	`str`	'generate_lsystem'
`technique`	`str`	''
`modality`	`str`	'image'
`params`	`dict`	Field(default_factory=dict)
`seed`	`int`	-1
`width`	`int`	256
`height`	`int`	256
`iterations`	`int`	5
`rule_string`	`str`	''
`axiom`	`str`	''
`fractal_type`	`str`	'mandelbrot'
`grammar_rules`	`dict`	Field(default_factory=dict)
`fitness_fn`	`str`	''
`content_base64`	`str`	''
`style_base64`	`str`	''
`preset_name`	`str`	''
`genome_id`	`str`	''
`artwork_id`	`str`	''
`palette`	`str`	'viridis'
`palette_stops`	`list[str]`	Field(default_factory=list)
`output_path`	`str`	''
`output_format`	`str`	''
`quality`	`str`	'medium'
`scene_type`	`str`	''
`expression`	`str`	''
`duration`	`float`	3.0
`attractor_type`	`str`	'clifford'
`feed_rate`	`float`	0.055
`kill_rate`	`float`	0.062
`particle_count`	`int`	5000
`cppn_layers`	`int`	3
`compose_with`	`list[dict]`	Field(default_factory=list)
`post_process`	`list[dict]`	Field(default_factory=list)
`tone_map`	`str`	'none'
`content_weight`	`float`	1.0
`style_weight`	`float`	1000000.0
`lloyd_iterations`	`int`	0
`backend`	`str`	'auto'
`goal_context`	`str`	''
`brief`	`str`	''
`custom_code`	`str`	''
`title`	`str`	''
`data`	`list[float]`	Field(default_factory=list)
`headless_enabled`	`bool`	False

`GenerativeOutput(BaseModel)`¶

Output from AdaptGenerativeArtBlock.

Field	Type	Default
`op`	`str`	required
`data_base64`	`str`	''
`output_path`	`str`	''
`modality`	`str`	''
`technique`	`str`	''
`width`	`int`	0
`height`	`int`	0
`seed_used`	`int`	0
`iterations`	`int`	0
`metadata`	`dict`	Field(default_factory=dict)
`svg_text`	`str`	''
`obj_text`	`str`	''
`text_output`	`str`	''
`backend`	`str`	''
`file_format`	`str`	''
`palette_used`	`str`	''
`frame_count`	`int`	0
`layers_composed`	`int`	0
`filters_applied`	`list[str]`	Field(default_factory=list)
`degraded`	`bool`	False
`degradation_reason`	`str`	''
`html_bundle`	`str`	''

`AdaptGenerativeArtMCPBlock(AIBlock[MCPGenerativeArtInput, MCPGenerativeArtOutput, dict])`¶

45-op generative art MCP block with SQLite persistence.

Field	Type	Default
`name`	`str`	'adapt_generative_art_mcp'
`state`	`dict`	field(default_factory=dict)
`db_path`	`str`	field(default_factory=lambda: os.environ.get('GENERATIVE_ART_DB_PATH', _DEFAULT_DB))
`resource_bounds`	`ResourceBounds`	field(default_factory=ResourceBounds)
`usage`	`ResourceUsage`	field(default_factory=ResourceUsage)

Methods:

`infer(data: MCPGenerativeArtInput) -> Result[MCPGenerativeArtOutput]`¶

`MCPGenerativeArtInput(BaseModel)`¶

Input to AdaptGenerativeArtMCPBlock — 50 ops.

Field	Type	Default
`op`	Literal['generate_lsystem', 'generate_ca', 'generate_fractal', 'generate_grammar', 'generate_evolutionary', 'generate_noise', 'generate_random_walk', 'generate_style_transfer', 'set_modality', 'set_params', 'set_seed', 'artwork_save', 'artwork_get', 'artwork_list', 'artwork_delete', 'genome_save', 'genome_get', 'genome_list', 'grammar_store', 'grammar_list', 'preset_store', 'preset_list', 'preset_apply', 'set_palette', 'set_quality', 'generate_manim', 'palette_list', 'palette_preview', 'export_artwork', 'generate_reaction_diffusion', 'generate_attractor', 'generate_voronoi', 'generate_flow_field', 'generate_cppn', 'generate_domain_coloring', 'generate_neural_style', 'generate_shape_grammar', 'compose_layers', 'apply_filters', 'stats', 'get_info', 'op_inventory', 'artwork_preview', 'artwork_open', 'list_patterns', 'backend_capabilities', 'backend_validate_source', 'backend_render_source', 'backend_preview_source', 'backend_export_source']	required
`technique`	`str`	''
`modality`	`str`	'image'
`params`	`dict`	Field(default_factory=dict)
`seed`	`int`	-1
`width`	`int`	256
`height`	`int`	256
`iterations`	`int`	5
`rule_string`	`str`	''
`axiom`	`str`	''
`fractal_type`	`str`	'mandelbrot'
`grammar_rules`	`dict`	Field(default_factory=dict)
`fitness_fn`	`str`	''
`content_base64`	`str`	''
`style_base64`	`str`	''
`artwork_id`	`str`	''
`genome_id`	`str`	''
`preset_name`	`str`	''
`name`	`str`	''
`tags_csv`	`str`	''
`description`	`str`	''
`limit`	`int`	50
`data_base64`	`str`	''
`genome_data`	`list[float]`	Field(default_factory=list)
`genome_fitness`	`float`	0.0
`rules_json`	`str`	''
`palette`	`str`	''
`palette_stops`	`list[str]`	Field(default_factory=list)
`output_path`	`str`	''
`output_format`	`str`	''
`quality`	`str`	'medium'
`scene_type`	`str`	''
`expression`	`str`	''
`duration`	`float`	3.0
`attractor_type`	`str`	'clifford'
`feed_rate`	`float`	0.055
`kill_rate`	`float`	0.062
`particle_count`	`int`	5000
`cppn_layers`	`int`	3
`compose_with`	`list[dict]`	Field(default_factory=list)
`post_process_chain`	`list[dict]`	Field(default_factory=list)
`tone_map`	`str`	'none'
`content_weight`	`float`	1.0
`style_weight`	`float`	1000000.0
`lloyd_iterations`	`int`	0
`backend`	`str`	'auto'
`goal_context`	`str`	''
`brief`	`str`	''
`custom_code`	`str`	''
`title`	`str`	''
`data`	`list[float]`	Field(default_factory=list)
`headless_enabled`	`bool`	False
`html_bundle`	`str`	''
`svg_text`	`str`	''
`obj_text`	`str`	''
`text_output`	`str`	''
`source_kind`	`str`	'sketch'
`source_text`	`str`	''
`render_target`	`str`	'html'
`timeout_ms`	`int`	5000
`allowed_features`	`list[str]`	Field(default_factory=list)
`resource_limits`	`dict`	Field(default_factory=dict)

`MCPGenerativeArtOutput(BaseModel)`¶

Output from AdaptGenerativeArtMCPBlock.

Field	Type	Default
`op`	`str`	required
`data_base64`	`str`	''
`modality`	`str`	''
`technique`	`str`	''
`width`	`int`	0
`height`	`int`	0
`seed_used`	`int`	0
`iterations`	`int`	0
`metadata`	`dict`	Field(default_factory=dict)
`svg_text`	`str`	''
`obj_text`	`str`	''
`text_output`	`str`	''
`backend`	`str`	''
`artwork_id`	`str`	''
`genome_id`	`str`	''
`name`	`str`	''
`found`	`bool`	False
`count`	`int`	0
`records`	`list[dict]`	Field(default_factory=list)
`message`	`str`	''
`degraded`	`bool`	False
`degradation_reason`	`str`	''
`output_path`	`str`	''
`file_format`	`str`	''
`palette_used`	`str`	''
`frame_count`	`int`	0
`layers_composed`	`int`	0
`filters_applied`	`list[str]`	Field(default_factory=list)
`html_bundle`	`str`	''
`declared_ops`	`list[str]`	Field(default_factory=list)
`handler_ops`	`list[str]`	Field(default_factory=list)
`ops_drift`	`list[str]`	Field(default_factory=list)
`completion_state`	`str`	'qualified-draft'
`warning_card`	`dict \\| None`	None
`evidence`	`list[str]`	Field(default_factory=list)
`capability_used`	`str`	''
`validation_status`	`str`	''
`dependency_status`	`dict`	Field(default_factory=dict)
`diagnostics`	`list[dict]`	Field(default_factory=list)
`source_kind`	`str`	''
`render_target`	`str`	''

`GenerativeArtStore`¶

SQLite-backed store for the adapt_generative_art MCP sub-package.

Constructor:

Parameter	Type	Default
`db_path`	`str`	':memory:'

Methods:

`save_artwork(name: str = '', technique: str = '', modality: str = '', data_base64: str = '', svg_text: str = '', obj_text: str = '', text_output: str = '', html_bundle: str = '', width: int = 0, height: int = 0, seed_used: int = 0, iterations: int = 0, metadata: dict | None = None, tags: str = '') -> str`¶

Save an artwork. Returns the assigned id.

`get_artwork(artwork_id: str) -> dict | None`¶

Retrieve an artwork by id.

`list_artworks(tags_csv: str = '', limit: int = 50) -> list[dict]`¶

List stored artworks, optionally filtered by tags.

`delete_artwork(artwork_id: str) -> bool`¶

Delete an artwork by id. Returns True if found.

`save_genome(name: str = '', genome: list[float] | None = None, fitness: float = 0.0, technique: str = '', tags: str = '', metadata: dict | None = None) -> str`¶

Save a genome. Returns the assigned id.

`get_genome(genome_id: str) -> dict | None`¶

Retrieve a genome by id.

`list_genomes(limit: int = 50) -> list[dict]`¶

List stored genomes.

`store_grammar(name: str, rules: dict, description: str = '', tags: str = '') -> str`¶

Store a grammar rule set. Returns the assigned id.

`list_grammars(limit: int = 50) -> list[dict]`¶

List stored grammars.

`get_grammar(name: str) -> dict | None`¶

Retrieve a grammar by name.

`store_preset(name: str, technique: str = '', params: dict | None = None, description: str = '', tags: str = '') -> str`¶

Store a parameter preset. Returns the assigned id.

`get_preset(name: str) -> dict | None`¶

Retrieve a preset by name.

`list_presets(limit: int = 50) -> list[dict]`¶

List stored presets.

`count_all() -> dict[str, int]`¶

Functions¶

`agentic_planner_enabled(default_enabled: bool = True) -> bool`¶

Decide whether the agentic (LLM) translate path should be used.

`validate_art_axes(axes: Any) -> dict[str, Any]`¶

Fail-safe-normalise an axes blob (e.g. an LLM fallback result).

`validate_translation(params: ArtParameters) -> ArtParameters`¶

Assert the resolved ArtParameters carries only in-vocabulary controlled axes.

MCP Tools¶

Operation	Source
`generate_lsystem`	generative_art_mcp
`generate_ca`	generative_art_mcp
`generate_fractal`	generative_art_mcp
`generate_grammar`	generative_art_mcp
`generate_evolutionary`	generative_art_mcp
`generate_noise`	generative_art_mcp
`generate_random_walk`	generative_art_mcp
`generate_style_transfer`	generative_art_mcp
`set_modality`	generative_art_mcp
`set_params`	generative_art_mcp
`set_seed`	generative_art_mcp
`artwork_save`	generative_art_mcp
`artwork_get`	generative_art_mcp
`artwork_list`	generative_art_mcp
`artwork_delete`	generative_art_mcp
`genome_save`	generative_art_mcp
`genome_get`	generative_art_mcp
`genome_list`	generative_art_mcp
`grammar_store`	generative_art_mcp
`grammar_list`	generative_art_mcp
`preset_store`	generative_art_mcp
`preset_list`	generative_art_mcp
`preset_apply`	generative_art_mcp
`set_palette`	generative_art_mcp
`set_quality`	generative_art_mcp
`generate_manim`	generative_art_mcp
`palette_list`	generative_art_mcp
`palette_preview`	generative_art_mcp
`export_artwork`	generative_art_mcp
`generate_reaction_diffusion`	generative_art_mcp
`generate_attractor`	generative_art_mcp
`generate_voronoi`	generative_art_mcp
`generate_flow_field`	generative_art_mcp
`generate_cppn`	generative_art_mcp
`generate_domain_coloring`	generative_art_mcp
`generate_neural_style`	generative_art_mcp
`generate_shape_grammar`	generative_art_mcp
`compose_layers`	generative_art_mcp
`apply_filters`	generative_art_mcp
`stats`	generative_art_mcp
`get_info`	generative_art_mcp
`op_inventory`	generative_art_mcp
`artwork_preview`	generative_art_mcp
`artwork_open`	generative_art_mcp
`list_patterns`	generative_art_mcp
`backend_capabilities`	generative_art_mcp
`backend_validate_source`	generative_art_mcp
`backend_render_source`	generative_art_mcp
`backend_preview_source`	generative_art_mcp
`backend_export_source`	generative_art_mcp