blender-pipeline-automation — Claude Code skill: features and install
The blender-pipeline-automation Claude Code skill automates batch work with Blender files and libraries. It handles folder-based conversion, inventory reports, and script runs.
MIT 509 installs ★ 2,624 updated 2026-11-13
npx skills add EL4CTEO/blender-skills --skill blender-pipeline-automation -a claude-codeWhat blender-pipeline-automation does
The skill describes running Blender in the background to convert .blend, .fbx, .glb, .obj, and .usd files, inventory folders, or run scripts across a collection. Jobs emit JSON results and failures; --shard splits work across processes.
It also covers loading and linking libraries, finding external paths, relocating files, and clearing unused data. The guide includes using bpy as a Python module and a GitHub Actions CI recipe. bpy needs Python 3.13 matched to the Blender release. Cycles CUDA/OptiX in containers needs the NVIDIA container runtime; headless EEVEE still needs GPU/EGL.
What it helps with
- convert a folder of
.fbxassets to.glb - generate a JSON inventory of a
.blendlibrary - run a QA script across a set of Blender files
- split batch work across processes
- relocate external paths in Blender files
Example prompts for the agent
- Convert the files in
assets/**/*.fbxtoglband save them inexport/. - Create a JSON inventory of the files in
library/. - Run
qa_check.pyonsrc/*.blendusing thegameprofile.
How to install blender-pipeline-automation
1Skills CLI (Vercel)
npx skills add EL4CTEO/blender-skills --skill blender-pipeline-automation -a claude-code2Manually: put the files in place
~/.claude/skills/blender-pipeline-automation/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/blender-pipeline-automation/
1Skills CLI (Vercel)
npx skills add EL4CTEO/blender-skills --skill blender-pipeline-automation -a codex2Manually: put the files in place
.agents/skills/blender-pipeline-automation/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add EL4CTEO/blender-skills --skill blender-pipeline-automation -a cursor2Manually: put the files in place
.cursor/skills/blender-pipeline-automation/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install EL4CTEO/blender-skills skills/blender-pipeline-automation2Manually: put the files in place
.github/skills/blender-pipeline-automation/SKILL.md1Skills CLI (Vercel)
npx skills add EL4CTEO/blender-skills --skill blender-pipeline-automation -a gemini-cli2Platform CLI
gemini skills install https://github.com/EL4CTEO/blender-skills --path skills/blender-pipeline-automationManually: put the files in place
.devin/skills/blender-pipeline-automation/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add EL4CTEO/blender-skills --skill blender-pipeline-automation -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about blender-pipeline-automation
How do I install blender-pipeline-automation in Claude Code?
Run npx skills add EL4CTEO/blender-skills --skill blender-pipeline-automation -a claude-code. The skill file is at ~/.claude/skills/blender-pipeline-automation/SKILL.md.
What does blender-pipeline-automation do?
It guides batch conversion, file inventory, and script execution for Blender assets. It also covers library handling, path management, and CI.
What does blender-pipeline-automation need to run?
Using bpy as a Python module requires Python 3.13 matched to the target Blender release. Cycles CUDA/OptiX in containers requires the NVIDIA container runtime, while headless EEVEE needs GPU/EGL.
Source text
SKILL.md · MIT
---
name: blender-pipeline-automation
description: Automate Blender 5.3 at scale - headless CLI flags, batch convert/inventory/run-any-skill-script over folders of .blend/.fbx/.glb/.obj/.usd, sharding, bpy pip module, library linking, path management, render-farm and CI patterns. Use for batch jobs, format conversion, library maintenance and pipeline tooling.
license: MIT
metadata:
author: blender-agent-skills
version: "1.2.0"
---
# Pipeline Automation
The execution model is in `blender-core` (`run_in_blender.py` finds Blender and prints only the JSON results).
## Batch one-liners
```
blender -b -P scripts/batch_process.py -- --in "assets/**/*.fbx" --op convert --to glb --out export/
blender -b -P scripts/batch_process.py -- --in library/ --op report # inventory JSON lines
blender -b -P scripts/batch_process.py -- --in "src/*.blend" --op script \
--script ../blender-asset-qa/scripts/qa_check.py --kwargs '{"profile":"game"}' # QA a whole library
blender -b -P scripts/batch_process.py -- --in src/ --op script --script ../blender-game-optimization/scripts/make_lods.py \
--kwargs '{"rename":true}' --save-to processed/ # LODs for every file
```
Each file is streamed as one JSON line with `ok`, timing and results. Failures don't stop the batch, and the summary lists them. **Parallel**: start N processes with `--shard 0/N` … `--shard N-1/N` (Blender uses one core for Python, so processes scale with CPU/GPU count).
## Headless CLI essentials
| Flag | Meaning |
|---|---|
| `-b` | background, no UI. Put it **first** |
| `file.blend` | open before running scripts. Order matters: args run left to right |
| `-P script.py` / `--python-expr "..."` | run Python |
| `--python-exit-code 1` | non-zero exit if the script raises (CI) |
| `--factory-startup` | ignore user prefs/add-ons (reproducible). Bundled add-ons still load |
| `--addons rigify,io_scene_fbx` | enable specific add-ons |
| `-E CYCLES -o //out/#### -F PNG -s 1 -e 10 -a` / `-f 5` | render (`blender-rendering`) |
| `--command extension build/validate/install-file` | extension tooling (`blender-addon-development`) |
| `-- --my args` | everything after `--` goes to the script (`sys.argv[sys.argv.index("--")+1:]`) |
| `--log "*"`, `--debug-python` | debugging. 5.0 changed the log format, so don't regex progress lines |
Environment variables: `BLENDER_USER_SCRIPTS`, `BLENDER_USER_EXTENSIONS`, `BLENDER_SYSTEM_PYTHON`. Use `OCIO` for custom color configs.
## bpy as a Python module
`pip install bpy==5.3.*` (or `5.2.*` for the LTS) (Python 3.13, matching the Blender release) gives `import bpy` in plain Python, which is handy for pytest/CI/web services. It has no UI and no GPU module by default, and it's large (~300 MB). Version-lock it with the Blender you target. The `bpy` wheel and the CLI behave the same for data API work.
## Library patterns
- **Append/link**: `bpy.data.libraries.load(path, link=...)` (see `blender-core/references/bpy-patterns.md`). 5.2: `data_from.libraries` exposes nested library paths when loading.
- **Write a subset**: `bpy.data.libraries.write(path, {datablocks}, fake_user=True, path_remap="RELATIVE_ALL")` builds asset/library files from a session.
- **Find files**: `bpy.data.file_path_foreach(...)` (5.2 flags: `EXPAND_TOKENS` for UDIMs, `EXPAND_SEQUENCES`, `EXPAND_CACHES`), and `bpy.utils.blend_paths(absolute=True)` lists every external path.
- **Remap/relocate**: `bpy.ops.file.find_missing_files(directory=...)`, `make_paths_relative`, `pack_all`. Libraries: `lib.filepath = new; lib.reload()`.
- **Iterate every ID**: `bpy.data.all_ids` (5.2; order isn't stable in 5.3, so sort by name if output must be deterministic), and `bpy.data.orphans_purge(do_recursive=True)` for cleanup.
## Robustness rules
- One task per process where possible. Long Blender sessions leak (images, undo). `batch_process.py` reuses the process but resets with `read_factory_settings(use_empty=True)` / `open_mainfile`.
- Never write into the source folder by default. Use separate `--out`/`--save-to`, and write with `copy=True`.
- Make scripts idempotent: re-running replaces their own outputs (named prefixes, `exist_ok`).
- Log JSON, not prose: one object per file/job, so agents and CI can parse it cheaply.
- Pin the Blender version (5.2 LTS, or 5.3) for a project. Record `bpy.app.version_string` + hash (`bpy.app.build_hash`) in the outputs.
- Paths: `bpy.path.abspath("//x")` is relative to the **opened** .blend. In headless batches use absolute paths, or set them after `open_mainfile`.
- GPU in containers: Cycles CUDA/OptiX needs the NVIDIA container runtime. Otherwise use CPU. EEVEE needs a GPU/EGL even headless.
## CI recipe (GitHub Actions)
Download the Linux tarball (`https://download.blender.org/release/Blender5.3/blender-5.3.0-linux-x64.tar.xz`), cache it, then `blender -b --factory-startup --python-exit-code 1 -P tests/run_all.py`. This repo's `.github/workflows/ci.yml` is a working example.
© blender-agent-skills · MIT · EL4CTEO/blender-skills