MCPInstall

blockbench-use

Load before Blockbench MCP calls that create, modify, or export content. Discover the active format and tools, route to modeling, UV/texturing, animation, particle, PBR, or Hytale guidance, and verify the requested result.

Installs
905
Last 24h
+9
Trending rank
#9317
Security
Not audited

Install

Works with every agent the skills CLI supports.

npx skills add https://github.com/jasonjgardner/blockbench-mcp-project --skill blockbench-use
  1. 01

    Install

    Run the command in your project. The CLI asks which agents to install for.

  2. 02

    Describe the task

    Ask your agent for something this skill covers. It loads blockbench-use automatically when relevant.

  3. 03

    Review the result

    The agent follows the SKILL.md steps, scripts and references bundled with the skill.

SKILL.md

The instructions your agent reads when the skill is triggered.

Blockbench Use

Load this skill before creating, modifying, or exporting Blockbench content. For capability questions, use MCP overview; plugin development uses blockbench-plugins.

Discover and Route

  1. Call get_capabilities: include_tools=true. It works without an open project and returns host/plugin identity, project (or null), registered formats, and current tool enabled states. Discover the running bundle rather than assuming every tool described in these source skills is installed.
  2. Preserve the intended project. Use its project.format_id; for a new project, choose an exact returned format ID for the user's target. get_capabilities: format_id="<ID>" inspects a format without switching projects. Generic Model is normally free, not generic. Hytale requires its registered Hytale format, not Bedrock. Read format and delivery guidance when choosing a target or preparing an export.
  3. Load relevant domain skills using the available skill-reading mechanism. Load additional guidance when the workflow reaches that domain; a simple edit does not need every skill.
  4. Inspect affected content with list_outline, list_textures, and targeted queries. Prefer UUIDs when names overlap. Inspect runtime mesh component keys with get_mesh_info before component edits.
  5. Check mode/selection when required. list_modes and set_mode navigate supported editor modes. Enabled tools can still require a compatible format, target, or selection. A feature value of null means unknown; compact supported_features lists true flags and unknown_features lists unknown flags.
IntentGuidance
Cubes, meshes, groups, silhouette, topologyModeling
Bedrock poles, cylinders, tubes, rings, and faceted round shapesModeling, then Bedrock primitives
UVs, pixel density, painting, layersTexturing
Vanilla/default Minecraft textures, or a named Minecraft block or material (bricks, oak logs, stone bricks)Vanilla textures, then Texturing for UVs
Animated textures, sprite sheets, vertical flipbooksFlipbook textures, which requires GPT Image textures for generated artwork
Pivots, keyframes, timing, animation exportAnimation
Particle effects, VFX, locators, particle keyframes, Snowstorm/Bedrock particle filesParticles, with Animation for the clip
Normal/height/MER materialsPBR materials
Normal, height or MER maps derived from an existing color textureAlbedo to normal, then PBR materials to assign them
Hytale formats, attachments, stretch, visibilityHytale, then applicable shared domains
Armature deformation, display slots, Bedrock material instancesMCP overview and live schemas

Work at the Scale of the Request

Before creating a new model or substantially redesigning its geometry, ask where the user's preference lies: appearance/accuracy, performance (fewer elements/faces), or a balance. Ask for the destination when unknown, including Minecraft edition and block versus entity. Reuse an explicit preference already given; "high quality" or "HD textures" alone does not choose a geometry budget. Continue discovery and inspection while awaiting the answer, but settle the preference before detailed geometry or texture generation. Routine edits that preserve an established design do not need this question again.

Use appearance and performance planning to translate the answer into a working budget, choose geometry versus texture/PBR detail, and measure the result. Read it before multiplying repeated parts, optimizing a slow model, or adding hardware details such as screws, bolts, rivets, and other fasteners, which normally belong in albedo and supported PBR channels rather than in geometry. For Bedrock custom blocks using multiple materials or cutout panels, also read material instances and texture delivery.

For a new model, also answer the real-time planning questions in real-time asset planning and record them in its plan template: target device, shading model, special materials, triangle budget, texel density and texture size, tiled versus atlas UVs, and animation type. Settle the shading model and texel density during blockout, because changing them later invalidates UVs and textures.

For a new asset, establish proportions and silhouette before detailed geometry, UVs, texture detail, and final animation. For an existing asset, inspect and change the requested area without rebuilding successful work. Use reference images to identify shape, palette, material and intended viewing distance. Treat Minecraft and Hytale art direction as target-specific guidance, not universal restrictions on every Blockbench format.

Before assigning detailed materials or repeating mapped geometry, follow UV scale and distortion guidance. Preserve face proportions and consistent material feature scale across different face sizes; do not stretch the same full atlas swatch over every face by default. Allow deliberate exceptions for uniform materials or designed effects, considering all PBR channels.

Choose verification that observes the changed behavior: inspect UV values, effective pixels per model unit in both directions, and a mapped checker for UV edits; inspect a texture image and the rendered model after painting; preview several times and the loop seam for animation. A screenshot of the rest pose alone cannot verify a walk cycle. See the delivery reference for format-specific checks.

Recovery and Delivery

  • Use save_checkpoint before exploratory or substantial edits when a history marker helps recovery. It is an undo-history marker, not a saved .bbmodel or a snapshot of unrecorded changes. It can clear a redo branch because it adds a history entry. Inspect get_undo_stack before recovery; count actual history entries rather than assuming one entry per tool call. Preserve intervening user edits.
  • Use dedicated undo / redo tools instead of triggering generic undo actions. Direct scripts must create their own correct native undo transaction. A marker cannot make an unrecorded edit reversible.
  • Prefer dedicated tools. If the task requires a feature absent from the live tools, use a supported native UI workflow or explain the specific limitation. Follow existing authorization and repository instructions for risky_eval; do not invent an unavailable wrapper or silently substitute a lossy format.
  • For a requested file, discover list_export_formats, choose an available codec with compile support, and use export_model. Save an editable .bbmodel alongside a runtime export when the requested handoff calls for an editable source. A compiled model does not establish that textures, animations, materials, controllers, or engine configuration have also been delivered.
  • Inspect returned export metadata and verify the actual destination/content. Truncated response text is a preview, not a complete file. When target-app access is unavailable, report the validation completed and the remaining integration check accurately.

FAQ

What does the blockbench-use skill do?

Load before Blockbench MCP calls that create, modify, or export content. Discover the active format and tools, route to modeling, UV/texturing, animation, particle, PBR, or Hytale guidance, and verify the requested result.

How do I install blockbench-use?

Run npx skills add https://github.com/jasonjgardner/blockbench-mcp-project --skill blockbench-use in your project. The skills CLI detects the agents you use and copies the skill into the right folder. For Claude Code you can also place the skill folder at ~/.claude/skills/blockbench-use/ (personal) or .claude/skills/blockbench-use/ (project).

Which AI agents can use blockbench-use?

Any agent that supports the open Agent Skills format (a folder with a SKILL.md file), including Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, Windsurf, Cline, OpenCode and more.

Is blockbench-use safe to install?

No partner security audit is available yet. Skills can run commands and scripts, so read the SKILL.md and any bundled files before installing.

Who maintains blockbench-use?

It is published by jasonjgardner in jasonjgardner/blockbench-mcp-project. It has 905 recorded installs.

More from jasonjgardner/blockbench-mcp-project

View all →

blockbench-modeling

jasonjgardner/blockbench-mcp-project

Create and edit 3D models in Blockbench using MCP tools. Use when building geometry with cubes, creating meshes, placing spheres/cylinders, editing vertices, extruding faces, or organizing models with groups. Covers both cube-based Minecraft modeling and freeform mesh editing.

AI & Agents1.3K installs

blockbench-texturing

jasonjgardner/blockbench-mcp-project

Create and paint textures in Blockbench using MCP tools. Use when creating textures, painting on models, using brush tools, filling colors, drawing shapes, applying gradients, managing texture layers, or working with UV mapping. Covers pixel art texturing, procedural painting, and UV manipulation.

AI & Agents1.2K installs

blockbench-animation

jasonjgardner/blockbench-mcp-project

Create and manage animations in Blockbench using MCP tools. Use when animating 3D models, creating keyframes, managing bone rigs, editing animation curves, or working with animation timelines. Covers walk cycles, idle animations, combat animations, and complex multi-bone animations.

Frontend & Design1.1K installs

blockbench-mcp-overview

jasonjgardner/blockbench-mcp-project

Discover and use Blockbench MCP tools, resources, and prompts. Use for capability questions, new-project planning, mode navigation, export, display transforms, Bedrock face materials, or native armature workflows.

AI & Agents1.1K installs

blockbench-pbr-materials

jasonjgardner/blockbench-mcp-project

Create and manage PBR materials in Blockbench using MCP tools. Use for Bedrock textureset.json import/export, normal or height maps, packed MER textures, material channel replacement, and uniform material properties.

AI & Agents861 installs

blockbench-hytale

jasonjgardner/blockbench-mcp-project

Create Hytale models and animations through Blockbench MCP. Use for character/attachment or prop formats, UV density, stretch, quads, shading, attachment pieces, and visibility keyframes. Requires the Hytale Models plugin.

AI & Agents836 installs

Popular ai & agents skills

View category →

google-agents-cli-publish

google/agents-cli

This skill should be used when the user wants to "publish an agent", "publish my ADK agent", "register an agent with Gemini", "publish to Gemini" (formerly Gemini Enterprise), or needs guidance on the agents-cli publish gemini command. Also use when the user wants to "manage agents in Agent Registry", "list/update/delete registered agents", or "register an MCP server". Covers ADK vs A2A registration modes, programmatic and interactive usage, flag reference, auto-detection from deployment metadata, Agent Registry fleet management, and troubleshooting. Part of the agents-cli skills suite. Do NOT use for deployment (use google-agents-cli-deploy).

AI & Agents513.9K installs

google-agents-cli-eval

google/agents-cli

This skill should be used when the user wants to "run an evaluation", "evaluate my agent", "evaluate my ADK agent", "write an eval dataset", "analyze eval failures", "compare eval results", "optimize agent", or needs guidance on the Gemini platform eval methodology and the Quality Flywheel. Covers eval metrics, dataset schema, LLM-as-judge scoring, and common failure causes. Applies to any agents-cli project, whatever framework the agent is written in. Do NOT use for agent API code patterns (ADK: use google-agents-cli-adk-code), deployment (use google-agents-cli-deploy), or project scaffolding (use google-agents-cli-scaffold).

AI & Agents513.5K installs

skill-creator

anthropics/skills

Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.

OfficialAI & Agents416.4K installs

codex-pet

prime-skills/runcomfy-agent-skills

Codex Pet generator on RunComfy. Build a Codex-compatible Codex Pet spritesheet.webp + pet.json from a single reference image, drop it into ${CODEXHOME:-$HOME/.codex}/pets/ / and Codex picks it up as a custom Codex Pet next to the 8 built-ins. This skill produces the exact Codex Pet atlas Codex expects (1536x1872 PNG/WebP, 8 cols x 9 rows, 192x208 cells, 9 animation states — idle, running-right, running-left, waving, jumping, failed, waiting, running, review). Calls OpenAI GPT Image 2 edit ONCE via the local RunComfy CLI as runcomfy run openai/gpt-image-2/edit to produce a canonical Codex Pet pose, then assembles all 9 animation rows programmatically with ImageMagick micro-transforms — no Codex Pro, no $imagegen, no OPENAIAPIKEY required, only RUNCOMFYTOKEN. Triggers on "codex pet", "create codex pet", "make codex pet", "hatch codex pet", "/hatch image", "desktop pet codex", "codex pets", "spritesheet.webp", or any explicit ask to build a custom pet for OpenAI Codex.

AI & Agents402.1K installs

full-output-enforcement

leonxlnx/taste-skill

Overrides default LLM truncation behavior. Enforces complete code generation, bans placeholder patterns, and handles token-limit splits cleanly. Apply to any task requiring exhaustive, unabridged output.

AI & Agents401K installs

executing-plans

obra/superpowers

Use when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available

AI & Agents245.7K installs