Blender addon management (compatibility alias — prefer blender_addons).

Operations mirror blender_addons: search, install_known, install_url, list_installed, enable, disable, info.

Return Format

Standard dict with keys: success, message, data

Examples

await call_tool("manage_blender_addons", {"operation": "search", "query": "gaussian splat"})

Manage Blender addons (portmanteau). Prefer this tool over manage_blender_addons.

Operations:

• list_addons / list_installed: List installed addons (requires Blender).

• install_addon: Install from local addon_path (path to .py or folder).

• install_from_url / install_url: Download and install from addon_url (ZIP or .py).

• install_known: Install from known registry by addon_name.

• uninstall_addon: Uninstall addon by addon_name.

• enable / disable: Enable or disable an installed addon module.

• search: Return known add-on URLs for a query (e.g. gaussian splat).

• info: Show known addons registry and addons directory.

Generate 3D meshes via external AI backends and import into Blender.

Backends (set API keys in environment):

• tripo: TRIPO_API_KEY

• rodin: RODIN_API_KEY or HYPER3D_API_KEY

• hunyuan: HUNYUAN3D_API_KEY (+ optional HUNYUAN3D_API_URL)

Operations:

• generate: text/image-to-3D, poll, download, import

• list_backends: show configured backends and env var names

Return Format

Standard dict with keys: success, message, data

Examples

await call_tool("blender_ai_generate", {"operation": "generate", "prompt": "a red chair"})

PORTMANTEAU PATTERN RATIONALE: Consolidates 17 related animation operations into single interface. Prevents tool explosion while maintaining full animation workflow functionality from basic keyframes to advanced character rigging. Follows FastMCP 2.14.3 best practices.

Comprehensive animation system for Blender supporting keyframes, shape keys, actions, constraints, and baking.

Animation Categories:

Basic Animation (7 operations):

• set_keyframe: Insert keyframes for location/rotation/scale at specific frames

• animate_location: Create movement animation between start/end frames

• animate_rotation: Create rotation animation with customizable curves

• animate_scale: Create scale animation with interpolation control

• play_animation: Start/stop viewport playback for preview

• set_frame_range: Define timeline start/end frames for scene

• clear_animation: Remove all keyframes from object (destructive)

Shape Keys (VRM facial expressions) (4 operations):

• list_shape_keys: Display all morph targets on mesh object

• set_shape_key: Set blend value (0.0-1.0) for shape key

• keyframe_shape_key: Insert keyframe for shape key animation

• create_shape_key: Create new shape key from current mesh state

Action Management (4 operations):

• list_actions: Show all animation actions in blend file

• create_action: Generate new action clip for object animation

• set_active_action: Assign action to object for playback

• push_to_nla: Push action to NLA track for layering/compositing

Interpolation & Timing (2 operations):

• set_interpolation: Set keyframe interpolation type (LINEAR, BEZIER, BOUNCE, ELASTIC, CONSTANT)

• set_easing: Configure easing curves (AUTO, EASE_IN, EASE_OUT, EASE_IN_OUT)

Constraints (2 operations):

• add_constraint: Add transform constraints to objects

• add_bone_constraint: Add pose constraints to armature bones

Baking for Export (2 operations):

• bake_action: Convert constraints to keyframes for export compatibility

• bake_all_actions: Consolidate NLA strips into single action

Look up Blender Python API documentation for an identifier.

Examples: Mesh, bpy.types.Mesh, bpy.ops.mesh.primitive_cube_add.

Use before writing bpy scripts to reduce hallucinated API calls.

Return Format

Standard dict with keys: success, message, data

Examples

await call_tool("blender_api_docs", {"identifier": "bpy.types.Object"})

PORTMANTEAU: Poly Haven + Sketchfab asset discovery and import (fleet subprocess Blender).

Operations:

• info: Capabilities and env vars (SKETCHFAB_API_TOKEN).

• polyhaven_categories: List category counts for hdris | textures | models.

• polyhaven_search: Search catalog (uses api.polyhaven.com/assets).

• polyhaven_import: Download + import — HDRIs set as world env; textures/models via files API bundles.

• sketchfab_search: Sketchfab v3 search (requires token).

• sketchfab_import: Download + import glTF from Sketchfab (requires token + downloadable model).

Advanced material and texture atlasing for VR performance optimization.

Reduces draw calls by intelligently merging materials and textures into atlas layouts, critical for mobile VR performance on Quest and similar devices.

Batch image and mesh export operations.

Operations:

• resize: resize images matching pattern (default *.png)

• convert: convert image formats in input_dir

• export: export mesh objects whose names contain name_pattern

Return Format

Standard dict with keys: success, message, data

Examples

await call_tool("blender_batch", {"operation": "resize", "input_dir": "C:/images"})

Create and control cameras in Blender scenes.

Supports multiple operations through the operation parameter:

• create_camera: Create a new camera

• set_active_camera: Set the active camera

• set_camera_lens: Adjust camera lens and sensor settings

Chat with the Blender AI assistant. Provides natural-language help with Blender operations.

Uses the blender-expert SKILL.md as system context so the LLM knows available tools and workflows. Uses ctx.sample() when the host supports sampling; falls back to Ollama.

Compositor graph operations for post-processing.

Operations:

• enable: turn on compositor nodes with render layers -> composite wiring

• add_node: add a compositor node

• connect_nodes: connect two compositor nodes

• glow: add a glow/glare effect chain

Return Format

Standard dict with keys: success, message, data

Examples

await call_tool("blender_compositor", {"operation": "enable"})

Return current webapp config (LLM provider, URLs, selected model, etc.).

Return Format

Standard dict with keys: success, message, data

Examples

await call_tool("config_get")

Update webapp config. Pass keys: server_host, server_port, theme, auto_sync, notifications, llm.

Return Format

Standard dict with keys: success, message, data

Examples

await call_tool("config_set", {"theme": "dark"})

PORTMANTEAU PATTERN RATIONALE: Consolidates natural language 3D construction into single agentic interface. Prevents tool explosion while enabling infinite 3D creativity through LLM-generated Blender scripts. Follows FastMCP 2.14.3 best practices.

Universal 3D object construction using natural language and LLM-generated Blender scripts.

This revolutionary tool enables creation of any 3D object through natural language descriptions by leveraging FastMCP 2.14.3 sampling to request SOTA LLM generation of Blender Python code.

Agentic Workflow:

1. Analysis: Parse natural language description and scene context

2. Sampling Request: Ask MCP client (SOTA LLM) to generate Blender Python script

3. Code Generation: LLM creates production-ready Blender automation code

4. Validation: Security and syntax validation of generated code

5. Execution: Safe execution in Blender with error handling

6. Iteration: Request refinements if needed (up to max_iterations)

Supported Complexity Levels:

• simple: Basic primitives, basic transforms, simple materials

• standard: Complex meshes, modifiers, materials, basic animation

• complex: Advanced geometry, rigging, physics, complex materials/textures

Style Presets:

• realistic: Physically accurate materials, lighting, proportions

• stylized: Artistic interpretation, exaggerated features, cartoon-like

• lowpoly: Minimal geometry, optimized for performance

• scifi: Futuristic design, metallic materials, glowing effects

Download from URL and import, or get supported formats info (portmanteau).

Operations:

• download: Download from url and optionally import into scene (requires url)

• info: Supported file formats and usage examples

Export Blender scenes and objects to interchange and game-engine formats.

Operations:

• export_gltf / export_glb / export_fbx / export_obj / export_stl / export_usd / export_vrm

• export_unity / export_vrchat / export_unreal (platform presets)

Return Format

Standard dict with keys: success, message, data

Examples

await call_tool("blender_export", {"operation": "export_glb", "output_path": "C:/out/model.glb"})

Platform-specific export presets for VR avatar deployment.

Handles the critical differences between VR platforms (scale, format, bone limits) ensuring your avatars work correctly in VRChat, Resonite, and Unity environments.

Create furniture and complex objects in Blender.

Supports multiple operations through the operation parameter:

• create_chair: Create chairs (dining, office, armchair, etc.)

• create_table: Create tables (dining, coffee, desk, etc.)

• create_bed: Create beds (single, double, bunk, etc.)

• create_sofa: Create sofas and couches

• create_cabinet: Create cabinets and storage

• create_shelf: Create bookshelves and shelving

• create_desk: Create desks and workstations

• create_stool: Create stools and bar stools

• create_room: Create room with walls, floor, ceiling, windows, doors

Geometry Nodes graph operations (procedural modeling).

Operations:

• create_group: create or reuse a GeometryNodeTree

• add_node: add a node to the group

• connect_nodes: wire two nodes in the group

• assign_modifier: attach node group to an object as Geometry Nodes modifier

• add_input: expose a group input socket

• list_node_types: common GeometryNode types for agents

Return Format

Standard dict with keys: success, message, data

Examples

await call_tool("blender_geonodes", {"operation": "create_group", "group_name": "MyNodes"})

PORTMANTEAU RATIONALE: Consolidates 12 Grease Pencil 2D animation operations into a single tool.

Operations:

• create: Create a new Grease Pencil object with layer and initial frame

• draw_stroke: Draw strokes (LINE, BOX, CIRCLE, ARC, CURVE) with color/thickness

• convert: Convert Grease Pencil to MESH, CURVE, or new GP strokes

• set_material: Create and assign GP material with stroke/fill color

• set_layer: Create, reorder, lock, toggle visibility of GP layers

• animate_stroke: Keyframe stroke properties (location, rotation, scale) over time

• onion_skinning: Enable/disable onion skin with before/after frame count

• add_modifier: Apply GP modifiers (BUILD, NOISE, SIMPLIFY, SMOOTH)

• fill_region: Fill enclosed stroke regions with color

• interpolate: Generate in-between frames between two GP frames

• delete_strokes: Remove strokes from a GP frame by selection type

• list_layers: List all layers and frame info on a GP object

Get help, list tools, search, tool info, or categories (portmanteau).

Operations:

• help: Comprehensive help for tools/functions (use function_name, category, detail_level)

• list_tools: List all tools, optionally filtered by category

• search: Search tools by name or description (use query)

• tool_info: Detailed info for one tool (use tool_name)

• categories: List all categories with tool counts

List all tools organized by category as a Prefab card.

Return Format

Prefab card with Headings per category and Rows per tool

Examples

await call_tool("show_tools_app")

Import 3D files into Blender scenes.

Supports multiple operations through the operation parameter:

• import_[format]: Import files in any supported format (FBX, OBJ, GLTF, STL, PLY, etc.)

• import_cad: Import CAD files (STEP, STP, IGES, IGS) with automatic conversion

• link_asset: Link external assets without importing

CAD IMPORT FEATURES:

• Automatic STEP/STP conversion using Mayo, FreeCAD, or online tools

• Mesh quality control (low/medium/high)

• Scale adjustment for unit conversion

• Direct integration with robotics-mcp CAD converter

Import CAD files (STEP, IGES) into Blender with automatic conversion.

This function handles the complete pipeline:

1. Detect CAD file format

2. Convert to mesh format using appropriate tool

3. Import converted mesh into Blender

Async job queue for long-running Blender Python scripts (renders, bakes, sims).

Operations:

• submit: queue script, returns job_id immediately

• status: poll job by job_id

• list: recent jobs

• cancel: cancel pending/running job

Return Format

Standard dict with keys: success, message, data

Examples

await call_tool("blender_jobs", {"operation": "submit", "script": "import bpy; print('hello')"})

PORTMANTEAU PATTERN RATIONALE: Consolidates 7 related lighting operations into single interface. Prevents tool explosion while maintaining full lighting workflow from basic lights to professional HDRI setups. Follows FastMCP 2.14.3 best practices.

Professional lighting system for Blender supporting all light types, HDRI environments, and lighting rigs.

Light Creation (4 operations):

• create_sun: Generate directional sunlight with shadow control for outdoor scenes

• create_point: Create omnidirectional point light for general illumination

• create_spot: Generate focused spotlight with beam angle and softness control

• create_area: Create rectangular area light for soft, realistic shadows

Lighting Setups (2 operations):

• setup_three_point: Create professional three-point lighting rig (key, fill, rim lights)

• setup_hdri: Configure HDRI environment lighting with world background

Light Management (1 operation):

• adjust_light: Modify properties of existing lights (energy, color, position, etc.)

Discover local LLM models from Ollama and LM Studio.

Return Format

Standard dict with keys: success, summary, result

Examples

await call_tool("list_local_models")

Generate a Blender Python script from a natural language prompt using a local LLM (Ollama).

Return Format

Standard dict with keys: success, script, error

Examples

await call_tool("generate_blender_script", {"prompt": "create a cube at origin", "model": "llama3.2"})

Portmanteau: list, pull, or remove Ollama models (CRUD for local LLM models).

Operations:

• list: return installed Ollama model names (and LM Studio if reachable).

• pull: pull model from Ollama registry (requires model_name). Slow for large models.

• remove: delete an Ollama model from disk (requires model_name).

View recent logs or log buffer statistics (portmanteau).

Operations:

• view: Recent log entries with optional filters (level_filter, module_filter, limit, since_minutes, include_details)

• stats: Log buffer statistics (total entries, time range, level distribution)

Show recent log entries as a scrollable Prefab card.

Return Format

Prefab card with timestamped log entries

Examples

await call_tool("show_logs_app", {"limit": 10})

Create and manage PBR materials in Blender.

Supports multiple operations through the operation parameter:

• create_fabric: Create fabric material (velvet, silk, cotton, linen, etc.)

• create_metal: Create metal material (gold, silver, brass, copper, etc.)

• create_wood: Create wood material (oak, pine, mahogany, walnut, etc.)

• create_glass: Create glass material (clear, tinted, frosted, stained)

• create_ceramic: Create ceramic material (porcelain, ceramic, terra_cotta)

• assign_to_object: Assign existing material to object

• create_from_preset: Create material from predefined preset

Convert non-standard (VRM/MToon) shaders to PBR textures.

Essential for VRM avatars moving into Resonite, Unity, or other PBR-based environments. Prevents "Semantic Dilution" of artistic intent.

PORTMANTEAU PATTERN RATIONALE: Consolidates 9 related mesh operations into single interface. Prevents tool explosion while maintaining full primitive creation and object management functionality. Follows FastMCP 2.14.3 best practices.

Create and manipulate mesh objects in Blender with comprehensive primitive support.

Core Operations:

• create_cube: Generate cube primitive with customizable dimensions

• create_sphere: Generate sphere primitive with vertex control

• create_cylinder: Generate cylinder primitive with radius/depth control

• create_cone: Generate cone primitive with radius/depth control

• create_plane: Generate plane primitive for ground/floor surfaces

• create_torus: Generate torus primitive with ring/tube radius control

• create_monkey: Generate Suzanne (monkey) primitive for testing

• duplicate_object: Create copies of existing objects with transforms

• delete_object: Remove objects from scene by name

• extrude: Extrude mesh region along +Z in edit mode

• inset: Inset selected faces

• bevel_modifier: Add bevel modifier to mesh object

• subdivide: Subdivide mesh edges

• merge_vertices: Merge by distance (remove doubles)

• delete_faces: Delete selected faces in edit mode

• join: Join multiple mesh objects into one

• separate_loose: Separate loose parts to new objects

• triangulate: Convert quads to triangles

Apply and manage mesh modifiers in Blender.

Supports multiple operations through the operation parameter:

• add_subsurf: Add subdivision surface modifier

• add_bevel: Add bevel modifier

• add_mirror: Add mirror modifier

• add_solidify: Add solidify modifier

• add_array: Add array modifier

• add_boolean: Add boolean modifier

• add_decimate: Add decimation modifier

• add_displace: Add displacement modifier

• add_wave: Add wave modifier

• remove_modifier: Remove a modifier

• apply_modifier: Apply modifier to mesh

• get_modifiers: List all modifiers on object

Create and manage particle systems and effects.

Supports multiple operations through the operation parameter:

• create_particle_system: Create basic particle system

• create_hair_particles: Create hair/fur particles

• create_fire_effect: Create fire/smoke particles

• create_water_effect: Create water/splash particles

• control_emission: Control particle emission settings

• bake_particles: Bake particle simulation

• set_particle_physics: Configure particle physics

Enable and configure physics simulations for objects.

Supports multiple operations through the operation parameter:

• enable_rigid_body: Add rigid body physics to object

• enable_cloth: Add cloth simulation to object

• enable_soft_body: Add soft body simulation to object

• enable_fluid: Add fluid simulation to object

• bake_physics: Bake physics simulation to keyframes

• add_force_field: Add force field to scene

• set_rigid_body_constraint: Add constraints between objects

• configure_world: Set up physics world settings

PORTMANTEAU PATTERN RATIONALE: Consolidates rendering operations into a single interface.

Render Operations:

• render_preview: Generate high-quality single frame preview

• render_turntable: Create 360-degree object rotation animation

• render_animation: Render full timeline animation sequence

• render_current_frame: Render only the current timeline frame

• screenshot_viewport: Capture viewport for agent vision (prefers live GUI session)

• render_multi_angle: Render N stills from orbit angles for review loops

• set_engine: Configure Cycles/EEVEE engine, samples, device

• configure_layers: Enable render passes on a view layer

• setup_post_processing: EEVEE bloom, SSAO, motion blur, DOF

Return Format

Standard string with operation result details

Examples

await call_tool("blender_render", {"operation": "render_preview", "output_path": "C:/render/output.png"})

Object repository management: save, load, search, list_objects.

• save: export Blender object to ~/.blender-mcp/repository/ with metadata

• load: append saved object into current scene with optional transforms

• search: filter index by query, category, tags, quality

• list_objects: return full index

Repository location: ~/.blender-mcp/repository/

Return Format

Standard dict with keys: success, message, objects

Examples

await call_tool("manage_object_repo", {"operation": "list_objects"})

AI-powered object construction and modification via sampling.

• construct: natural language → Blender Python → execute in scene

• construct_and_save: construct then immediately save to repository

• modify: find stored script for object, sample modification, validate, execute

Requires a client that supports MCP sampling (Claude Desktop, Antigravity, etc.)

Return Format

Standard dict with keys: success, message, object_name, scene_objects_created

Examples

await call_tool("manage_object_construction", {"operation": "construct", "description": "A red cube", "name": "MyCube"})

Export a repository asset with platform-specific optimisations for cross-MCP handoff.

Supported targets: vrchat (FBX), resonite (GLB). asset_id must exist in ~/.blender-mcp/repository/. Writes actual export files to a temp directory and returns their paths.

Return Format

Standard dict with keys: success, asset_id, target_mcp, primary_files, supporting_files

Examples

await call_tool("export_for_mcp_handoff", {"asset_id": "...", "target_mcp": "vrchat"})

PORTMANTEAU PATTERN RATIONALE: Consolidates 11 related rigging operations into single interface. Prevents tool explosion while maintaining full character rigging workflow from armature creation to humanoid mapping. Follows FastMCP 2.14.3 best practices.

Complete character rigging system for Blender supporting armatures, bones, IK, skinning, and humanoid standards.

Armature Operations (4 operations):

• create_armature: Generate new skeleton object with customizable positioning

• add_bone: Add individual bones to existing armature with parent/child relationships

• create_bone_ik: Set up inverse kinematics constraints for realistic joint movement

• create_basic_rig: Auto-generate complete biped character rig with standard bone structure

Bone Management (3 operations):

• list_bones: Display all bones in armature with hierarchy and properties

• pose_bone: Set bone transformations in pose mode for animation

• set_bone_keyframe: Insert keyframes for bone animation at specific frames

Pose & Animation (1 operation):

• reset_pose: Return armature to rest pose, clearing all pose transformations

Skinning & Weights (2 operations):

• transfer_weights: Copy vertex weights between meshes using various projection methods

• manage_vertex_groups: Create, rename, mirror, or remove vertex groups for weight painting

Standards & Compatibility (1 operation):

• humanoid_mapping: Apply VRChat/Unity humanoid bone naming and structure standards

Comprehensive scene management for Blender.

Supports multiple operations through the operation parameter:

• create_scene: Create a new scene

• list_scenes: List all scenes in the file

• clear_scene: Clear all objects from current scene

• set_active_scene: Set the active scene

• link_object_to_scene: Link object to a scene

• create_collection: Create a new collection

• add_to_collection: Add object to collection

• set_active_collection: Set the active collection

• set_view_layer: Set active view layer

• setup_lighting: Set up scene lighting

• setup_camera: Set up scene camera

• set_render_settings: Configure render resolution

Execute Python script in Blender (headless).

Sculpt mode operations for organic mesh editing.

Operations:

• enter: switch object to sculpt mode

• exit: leave sculpt mode (default back to OBJECT)

• set_brush: assign brush with strength and radius

• dynotopo: enable/disable dynamic topology

• symmetrize: mirror sculpt along an axis

• mask_clear: clear sculpt mask

• remesh_voxel: apply voxel remesh modifier

• list_brushes: common sculpt brush names

Return Format

Standard dict with keys: success, message, data

Examples

await call_tool("blender_sculpt", {"operation": "enter", "object_name": "Cube"})

Select objects and elements in Blender scenes.

Supports multiple operations through the operation parameter:

• select_objects: Select specific objects by name

• select_by_type: Select all objects of a specific type

• select_by_material: Select objects using a specific material

• select_all: Select all objects in scene

• select_none: Deselect all objects

• invert_selection: Invert current selection

Manage a Blender GUI session from the MCP server.

Operations:

• status: check whether a Blender GUI is running under MCP control

• start: launch Blender GUI (optionally opening a .blend file)

• stop: kill the managed Blender GUI process

• run_script: execute a Python snippet in the running Blender session (requires the bridge addon to be enabled in Blender)

• demo: run a named built-in demo in Blender via the bridge Available demos: living_room_with_car, driver_training, garden, house_interior

Shader node graph operations for materials.

Operations:

• create_material: create a node-based material

• create_node: add a shader node to an existing material

• connect_nodes: wire two shader nodes together

• list_node_types: return common ShaderNode type names

Return Format

Standard dict with keys: success, message, data

Examples

await call_tool("blender_shaders", {"operation": "create_material", "material_name": "MyMat"})

Comprehensive shape key management for facial animation and VRM avatars.

Handles viseme creation for lip sync, blink animations, facial expressions, and VRM compliance checking for VR platforms.

Advanced Gaussian Splatting (3DGS) management for hybrid environments.

Supports importing real-world captures, cleaning them up, and preparing them for VR platforms like Resonite for immersive experiences.

System status and monitoring (portmanteau).

Operations:

• status: MCP server, Blender, system, and performance summary

• system_info: Detailed OS, Python, env, and resources

• health_check: Blender availability, resources, tool registration

• performance_monitor: Sample CPU/memory/disk over duration_seconds (max 60)

Return Blender MCP server information, version, and status.

Return Format

JSON string with keys: name, version, status, blender_executable, blender_status, platform, python_version

Examples

await call_tool("server_info")

Show Blender MCP server status as a rich dashboard card.

Return Format

Prefab card with KPIs and system status

Examples

await call_tool("show_blender_status_app")

Create and manage textures in Blender.

Supports multiple operations through the operation parameter:

• create_[type]: Create procedural textures (noise, voronoi, musgrave, wave, checker, brick, gradient)

• assign_texture: Assign texture to material

• bake_texture: Bake textures from objects

Transform objects in 3D space.

Supports multiple operations through the operation parameter:

• set_location: Set object position

• set_rotation: Set object rotation (degrees)

• set_scale: Set object scale

• translate: Move object by offset

• rotate: Rotate object by angle

• scale: Scale object by factor

• apply_transform: Apply transforms to mesh

• reset_transform: Reset transforms to identity

Manage UV mapping and texture coordinates.

Supports multiple operations through the operation parameter:

• unwrap: Unwrap UV coordinates

• smart_project: Smart UV projection

• cube_project: Cube projection

• cylinder_project: Cylindrical projection

• sphere_project: Spherical projection

• reset_uvs: Reset UV coordinates

• get_uv_info: Get UV mapping information

Comprehensive validation tools for avatars and 3D models.

Performs pre-flight checks to ensure models meet platform requirements for VRChat, Resonite, Unity, and other 3D platforms.

Agent vision refinement loop: capture, review bundle, apply fixes.

Operations:

• capture: viewport PNG + base64 for vision models

• review_bundle: screenshot + multi-angle stills + scene summary + refinement prompt

• apply_script: run corrective bpy script after vision model feedback

Return Format

Standard dict with keys: success, message, data

Examples

await call_tool("blender_vision_refine", {"operation": "capture", "output_path": "C:/shot.png"})

Comprehensive VRM metadata management for avatar configuration.

Handles all VRM-specific settings required for proper avatar functionality in VRChat, VRoid Studio, and other VRM-compatible platforms.

PORTMANTEAU PATTERN RATIONALE: Consolidates 20 VSE operations into a single interface. Prevents tool explosion while providing full video editing: strip management, timeline editing, effects, and rendering. Follows FastMCP 3.x best practices.

Blender's built-in Video Sequence Editor (VSE) exposed as MCP tools. Add video/audio/image strips, apply transitions, trim, cut, move, and render to video.

Strip Creation (7 operations):

• add_movie: Add a video/MP4 file as a movie strip

• add_sound: Add an audio file (WAV, MP3, etc.) as a sound strip

• add_image_sequence: Add a folder of images as an image sequence strip

• add_scene: Add a 3D scene as a strip (compositing)

• add_color: Add a solid color matte strip

• add_text: Add a text overlay strip

• add_effect: Add a transition/filter effect between strips

Strip Editing (6 operations):

• delete_strip: Remove a strip by name

• cut_strip: Cut a strip at a specific frame

• trim_strip: Set strip start/end frames (trim handles)

• move_strip: Move a strip to a different channel or frame

• mute_strip: Mute or unmute a strip

• lock_strip: Lock or unlock a strip

Properties (3 operations):

• set_speed: Change playback speed (creates speed effect strip)

• set_blend: Set blend mode and opacity (ALPHA_OVER, CROSS, ADD, etc.)

• set_transform: Set position, scale, rotation of a strip

Information (2 operations):

• list_strips: List all strips in the timeline with properties

• get_timeline_info: Get timeline frame range, FPS, strip count, channels

Rendering (1 operation):

• render_video: Render the VSE timeline to a video file (H264/MPEG4)

Cleanup (1 operation):

• clear_vse: Remove all strips from the timeline

Execute multiple Blender operations in a single call (macro/batch).

This tool enables complex workflows without multiple round-trips to the MCP server. Operations are executed sequentially, with optional variable passing between steps.

OPERATIONS:

• list_templates: List available workflow templates

• get_template: Get details of a specific template

• execute: Execute a workflow (from steps or template)

STEP FORMAT: Each step is a dict with:

• tool: The blender tool name (e.g., "blender_mesh")

• operation: The operation within that tool

• ...other parameters for that operation

• as: (optional) Store result with this name for later steps

• if_result: (optional) Only run if previous result contains this string

VARIABLE REFERENCES: Use $varname to reference results from previous steps. Use ${varname.field} to reference specific fields from JSON results.

Create a new Blender scene with the specified name.

Return Format

Standard dict with keys: success, message, data

Examples

await call_tool("create_scene", {"scene_name": "MyScene"})

List all scenes in the current Blender file.

Return Format

Standard dict with keys: success, message, data

Examples

await call_tool("list_scenes")

Remove all objects from the current scene.

Return Format

Standard dict with keys: success, message, data

Examples

await call_tool("clear_scene")

Get the full scene hierarchy as a JSON-formatted string.

Return Format

Standard dict with keys: success, message, data

Examples

await call_tool("scene_get_hierarchy")

Set the active scene by name.

Return Format

Standard dict with keys: success, message, data

Examples

await call_tool("set_active_scene", {"scene_name": "Scene"})

Create a new collection.

Return Format

Standard dict with keys: success, message, data

Examples

await call_tool("create_collection", {"collection_name": "MyCollection"})

Add an object to a collection.

Return Format

Standard dict with keys: success, message, data

Examples

await call_tool("add_to_collection", {"collection_name": "MyCollection", "object_name": "Cube"})

Execute autonomous multi-step Blender workflows via FastMCP 3.1 SEP-1577 sampling.

The client LLM plans and executes a 3D workflow step-by-step, autonomously calling Blender capability probes to inform each decision, looping until the goal is achieved or max_steps is exhausted.

Intelligent batch 3D scene processing via FastMCP 3.1 SEP-1577 multi-step sampling.

The LLM autonomously queries material, modeling, and IO capabilities to build a processing pipeline tailored to each scene's needs.

Conversational Blender assistant with SEP-1577 multi-step sampling.

The LLM may probe capabilities to give accurate, operation-specific answers before responding. Falls back gracefully when sampling is not available.