Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add instructions/dwgx/blender-copilot/claude-mdgit clone --depth 1 https://github.com/dwgx/blender-copilotWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/instructions/dwgx/blender-copilot/claude-md)<a href="https://agentmods.dev/instructions/dwgx/blender-copilot/claude-md"><img src="https://agentmods.dev/badge/instructions/dwgx/blender-copilot/claude-md.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.02720 | $0.02720 |
| Opus 5 | $0.01360 | $0.01360 |
| Sonnet 5 | $0.00544 | $0.00544 |
| Haiku 4.5 | $0.00272 | $0.00272 |
Grade A, and why
blender-copilot CLAUDE.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 5d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 164 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md
This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
Blender Copilot — Agent Onboarding Guide
What This Project Is
An MCP (Model Context Protocol) server that lets AI agents control Blender for 3D modeling. 308 tools total — the most comprehensive Blender MCP server available. Full zero-to-published VRChat avatar pipeline plus complete 3D workflow automation.
The user (DWGX) wants the AI to operate like a real 3D modeler — not just execute commands, but understand mesh topology, weight painting, UV layout, VRChat avatar constraints, and make artistic/technical decisions autonomously.
AI Working Principles (Core — applies to ALL sessions)
The AI is a master executor, not a designer. This is how every interaction works:
- Aesthetics & design judgment → HUMAN. The AI never assumes what looks good. When aesthetic choices arise, ASK the user. Research references, present options with visual/technical tradeoffs, but the human makes the call.
- Form & shape description → HUMAN. The user describes the shape, silhouette, proportions, character feel. The AI's job is to understand the description as deeply as possible — ask clarifying questions about spatial relationships, curvature intent, muscle/fat distribution, stylization level.
- Sculpting → AI must be PERFECT. This is non-negotiable. The AI must:
- Master every Blender sculpt brush parameter (radius, strength, falloff, stroke method, texture)
- Execute systematic sculpting passes: primary forms → secondary forms → tertiary detail
- Understand anatomy (bone landmarks, muscle insertions, fat pads, skin tension)
- Use multi-resolution workflow correctly (base shape at low subdiv, detail at high)
- Generate the most precise sculpt code possible — exact coordinates, curves, pressure profiles
- Research and apply the best known techniques for each sculpting task
- Execution precision → AI must MAXIMIZE. For every operation (not just sculpting):
- Push parameter precision to the limit
- Understand the "why" behind each value, not just copy defaults
- When uncertain, research first (use codex-subagent for heavy lookups)
- Proactive questioning: When the user gives a vague shape description, don't guess — ask structured questions:
- "这个曲面是要偏硬（机械感）还是偏软（有机感）？"
- "参考图里的这个转折，你想要多锐利？"
- "这个比例是写实还是stylized？大概几头身？"
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 5d ago First seen · 164 lines · 2,720 tokens per session scan A 99b4d9fcfa13
blender-copilot CLAUDE.md is an instructions file published in the GitHub repository dwgx/blender-copilot (3 stars, last pushed 2mo ago), licensed MIT. It adds 2,720 tokens to every session, about $0.0136 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
zlibrary-mcp AGENTS.md
AGENTS.md instructions for rookslog/zlibrary-mcp, covering agents.md, what this is, setup and commands, invariants that break silently and architecture.
zlibrary-mcp CLAUDE.md
Claude Code instructions for rookslog/zlibrary-mcp, covering claude.md, claude code specifics, reading order for a cold start, worktrees and useful mcp servers here.
blender-open-mcp AGENTS.md
AGENTS.md instructions for dhakalnirajan/blender-open-mcp, covering blender open mcp — agents.md, project overview, repository layout, key design decisions and conventions.
ThreeBrowserStudio AGENTS.md
AGENTS.md instructions for SamG-Coder/ThreeBrowserStudio: ThreeBrowser Studio is an LLM-first authoring runtime. MCP is the editor and the native WebGPU window is the shared live viewport. Preserve that product shape: do not grow a conventional inspector-heavy browser UI.
draftly-wps CLAUDE.md
Claude Code instructions for Antares422/draftly-wps, covering claude.md, 项目是什么, 硬性规则, 部署流程 and 架构.
mcp-deadmansnitch CLAUDE.md
Instructions for jamesbrink/mcp-deadmansnitch, covering claude.md, dead man's snitch mcp server, essential commands, run tests (uses pytest-asyncio in auto mode) and type checking and linting.