About

Imported from 1999AZZAR/ll3m-mcp (skill/ll3m-agent/SKILL.md). Install upstream with npx skills add 1999AZZAR/ll3m-mcp --skill ll3m-agent. Copyright stays with the author.

LL3M Agent Skill (v4.2)

Name: ll3m-agent
Author: 1999AZZAR

Expert autonomous 3D modeling assistant for Blender, specialized in high-fidelity "Perfect Clone" generation using Recursive Iteration.

Recursive Modeling Workflow

Follow this loop until the model is physically and semantically perfect:

1. Initialization (Stage I)

PLAN: Architect components.
WRITE: Generate base topology and PBR materials.
EXECUTE: Create the initial scene.

2. The Auto-Iteration Loop (Stage II & IV)

CAPTURE: Use get_fast_feedback for rapid, low-compute verification.
CRITIQUE: Use the critic persona to generate a Quality Score and a Geometric Delta.
RECURSE:
- IF Quality Score < 90:
  1. Apply fixes using execute_staged_refinement.
  2. Update Blender HUD with helpers.draw_agent_hud(message="Refining: Iteration X").
  3. REPEAT the CAPTURE -> CRITIQUE step.
- IF Quality Score >= 90:
  1. Proceed to Finalization.

3. Finalization (Stage III)

HI-FI: Use get_modeling_patterns for final bevels and SubD.
RENDER: Execute render_output for the final beauty shot.

Technical Standards

Loop Limit: Max 3 iterations for minor objects, 5 for complex clones.
Physics: Mandatory use of helpers.run_gravity_settle() or helpers.snap_to_surface() during the loop.
Persistence: Always save_blend after a successful >=90 score.

Use it

Copy one of these into your project. Installing also returns the manifest and these snippets.

yaml

targets:
  - https://api.opensmartroute.ai/api/v1/registry/1999azzar-ll3m-mcp-ll3m-agent/manifest   # or paste the manifest below

Manifest

An Open Capability Manifest: the router reads it to know what this does, what it costs and when to pick it.

1999azzar-ll3m-mcp-ll3m-agent.ocm.jsonjson

{
  "ocm": "1",
  "id": "1999azzar-ll3m-mcp-ll3m-agent",
  "kind": "skill",
  "name": "ll3m-agent",
  "description": "Expert autonomous 3D modeling assistant for Blender, specialized in high-fidelity clone generation via recursive iteration.",
  "publisher": "1999AZZAR",
  "version": "1.0.0",
  "capabilities": {
    "domains": [
      "general"
    ],
    "tags": [
      "skill-md",
      "blender",
      "3d-modeling",
      "recursive-iteration",
      "autonomous",
      "github"
    ],
    "languages": [
      "en"
    ]
  },
  "quality_prior": 0.6,
  "examples": [
    "Expert autonomous 3D modeling assistant for Blender, specialized in high-fidelity clone generation via recursive iteration."
  ],
  "primary": false,
  "metadata": {
    "source": {
      "provider": "github",
      "repository": "https://github.com/1999AZZAR/ll3m-mcp",
      "path": "skill/ll3m-agent/SKILL.md",
      "ref": "e7163cfa9477d6ca2d1509f53fc4925a82e86168",
      "url": "https://github.com/1999AZZAR/ll3m-mcp/blob/e7163cfa9477d6ca2d1509f53fc4925a82e86168/skill/ll3m-agent/SKILL.md",
      "key": "1999AZZAR/ll3m-mcp/skill/ll3m-agent/SKILL.md"
    }
  },
  "instructions": "# LL3M Agent Skill (v4.2)\n\nExpert autonomous 3D modeling assistant for **Blender**, specialized in high-fidelity \"Perfect Clone\" generation using **Recursive Iteration**.\n\n## Recursive Modeling Workflow\n\nFollow this loop until the model is physically and semantically perfect:\n\n### 1. Initialization (Stage I)\n- **PLAN**: Architect components.\n- **WRITE**: Generate base topology and PBR materials.\n- **EXECUTE**: Create the initial scene.\n\n### 2. The Auto-Iteration Loop (Stage II & IV)\n- **CAPTURE**: Use `get_fast_feedback` for rapid, low-compute verification.\n- **CRITIQUE**: Use the `critic` per",
  "cost": {
    "context_tokens": 358
  }
}

Fetch it by URL: GET /api/v1/registry/1999azzar-ll3m-mcp-ll3m-agent/manifest?version=1.0.0

Reviews

Star ratings from people who tried it. One review per account; edit yours any time.

No reviews yet. Install it, try it, and be the first to rate it.