luxar-hpc-batch-fit SKILL.md skillA
luxar-hpc-batch-fit from royerlab/luxar is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.
mdr add royerlab/luxar/luxar-hpc-batch-fit@git:20261018.1c54257Pins git:20261018.1c54257 in mdr.lock. Pin the fingerprint instead.
- Audit grade
- A, 17 of 17 checks passed
- Kind
- SKILL.md skill
- Repository
- royerlab/luxar
- Path
.agents/skills/luxar-hpc-batch-fit/SKILL.md- Versions recorded
- 6
- Latest version
- git:20261018.1c54257, committed Oct 18, 2026
- First version recorded
- Jul 1, 2026
- Size
- 6,257 bytes
- Repository stars
- 75
- License
- BSD-3-Clause
Fit Gaussian splats to a whole nD timelapse at scale — across many GPUs on one box (local) or a Slurm/HPC cluster array job. Use when a user has a large multi-timepoint / multi-channel OME-Zarr volume and wants to fit it end to end: plan tiles over T×C, fit every tile, then stream-merge to one kind=partition .gsplats.zarr. Covers batch-fit run (local multi-GPU), batch-fit submit (Slurm), status/validate/merge/cancel, the gsplat benchmark GPU profile, and running on the CZ Biohub "Bruno" cluster.
Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.
What the file says
# Luxar HPC / batch Gaussian-splat fitting `batch-fit` fits a whole nD timelapse by planning tiles once over T×C, fitting each tile, then running a **memory-safe streaming merge** to one `kind=partition` `.gsplats.zarr`. Two entrypoints, same plan/merge machinery: - **`batch-fit run`** — LOCAL multi-GPU (one box, no Slurm). Saturates every GPU. - **`batch-fit submit`** — Slurm cluster array job. `status` / `validate` / `merge` / `cancel` are shared. Input must be **OME-Zarr-style** (it discovers T/C/spatial structure); flat `.npy`/`.tiff` is rejected — for a single small volume use plain `luxar gsplat fit` instead. ## Decide: local vs cluster | | `batch-fit run` (local) | `batch-fit submit` (Slurm) | | --- | --- | --- | | Where | one machine, N GPUs | HPC cluster array | | Key flag | `--gpus auto/all/cpu/0,1` | `--partition`/`-p` (REQUIRED) | | Resumable | yes (`--no-resume` to force) | yes (re-submit skips done tiles) | ## Local multi-GPU (no Slurm) ```bash # Uniform tiles, every GPU above a free-VRAM floor; resumable. luxar gsplat batch-fit run vol.zarr out/ --gpus auto --tile-size 256 # Content-balanced boxes (needs a density) + per-part LOD baked into the merge. …
Read the whole file at its exact version.
How to install
mdr add royerlab/luxar/luxar-hpc-batch-fit@sha256:4ad17358e37576d7The line at the top of this page pins the label git:20261018.1c54257, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.
[](https://markdownregistry.com/a/art_a6s6wuhd6kn7p7c6)
2 badge image fetches in 30 days. What the badge shows.
Versions
The registry has recorded 6 versions of this file, between Jul 1, 2026 and Oct 18, 2026, about one every 22 days.
| version | committed | commit | size | audit | |
|---|---|---|---|---|---|
| git:20261018.1c54257 latest | 2026-10-18 | 1c54257 | 6,257 B | A | view · diff |
| git:20260921.5af784f | 2026-09-21 | 5af784f | 5,670 B | A | view · diff |
| git:20260806.977be53 | 2026-08-06 | 977be53 | 5,680 B | A | view · diff |
| git:20260714.639a926 | 2026-07-14 | 639a926 | 5,670 B | A | view · diff |
| git:20260703.d5d99b0 | 2026-07-03 | d5d99b0 | 5,588 B | A | view · diff |
| git:20260701.1fea71e | 2026-07-01 | 1fea71e | 5,600 B | A | view |
Audit of the latest version
- pass: Frontmatter block present
- pass: Frontmatter declares a name
- pass: Frontmatter declares a description
- pass: Size between 200 bytes and 200 KB (6257 bytes)
- pass: No zero-width or bidi control characters
- pass: No instruction hidden inside an HTML comment
- pass: No link to an exfiltration or paste host
- pass: No credential-shaped string
- pass: No instruction to send local credentials anywhere
- pass: No text hidden with inline styles
- pass: No prompt-injection phrasing
- pass: No curl or wget piped into a shell
- pass: No recursive delete of root, home or parent
- pass: No instruction to read or print local credentials
- pass: No base64 blob over 200 characters
- pass: No link to a raw IP address
- pass: No script tag
Source
royerlab/luxar · 75 stars · license BSD-3-Clause · pushed 2026-10-21 · branch main
API
GET https://markdownregistry.com/api/v1/artifacts/art_a6s6wuhd6kn7p7c6 GET https://markdownregistry.com/api/v1/resolve?ref=royerlab/luxar/luxar-hpc-batch-fit GET https://markdownregistry.com/api/v1/blob/4ad17358e37576d7f3ef0df52e74c3b7d99dc48bc199b8d0c4deb4938e32dc28
Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.
Guides: the SKILL.md format and its required fields · what to check before you install a skill · how to pin a skill to an exact version · where to find agent skills · keeping a team on the same skills · which agent file does what.