Qwen Mtp Gguf
R6410418/Jackrong-llm-finetuning-guide
Complete agent-ready workflow for Qwen-family MTP or nextn GGUF conversion and release.
Develop, deploy, and maintain the DynaCell virtual-staining HuggingFace demo hosted at biohub/dynacell (ZeroGPU).
Install Hf Dynacell
$ npx skills add mehta-lab/VisCy --skill hf-dynacell -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install mehta-lab/VisCy hf-dynacell --agent claude-codeProject scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).
$ git clone --depth 1 https://github.com/mehta-lab/VisCy.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.claude/skills/hf-dynacell .claude/skills/hf-dynacell && rm -rf skills-srcUse ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.
Claude Code skills documentation · loads skills from .claude/skills/
Install the "hf-dynacell" agent skill from https://github.com/mehta-lab/VisCy/tree/main/.claude/skills/hf-dynacell into .claude/skills/hf-dynacell/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "hf-dynacell", then confirm the skill loads.Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$skill-installer install https://github.com/mehta-lab/VisCy/tree/main/.claude/skills/hf-dynacellType this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.
$ npx skills add mehta-lab/VisCy --skill hf-dynacell -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install mehta-lab/VisCy hf-dynacell --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/mehta-lab/VisCy.git skills-src && mkdir -p .agents/skills && cp -r skills-src/.claude/skills/hf-dynacell .agents/skills/hf-dynacell && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "hf-dynacell" agent skill from https://github.com/mehta-lab/VisCy/tree/main/.claude/skills/hf-dynacell into .agents/skills/hf-dynacell/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "hf-dynacell", then confirm the skill loads.Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add mehta-lab/VisCy --skill hf-dynacell -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install mehta-lab/VisCy hf-dynacell --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/mehta-lab/VisCy.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/.claude/skills/hf-dynacell .cursor/skills/hf-dynacell && rm -rf skills-srcUse ~/.cursor/skills/ instead of .cursor/skills for a personal install.
Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/
Install the "hf-dynacell" agent skill from https://github.com/mehta-lab/VisCy/tree/main/.claude/skills/hf-dynacell into .cursor/skills/hf-dynacell/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "hf-dynacell", then confirm the skill loads.Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gemini skills install https://github.com/mehta-lab/VisCy.git --path .claude/skills/hf-dynacell--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.
$ npx skills add mehta-lab/VisCy --skill hf-dynacell -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install mehta-lab/VisCy hf-dynacell --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/mehta-lab/VisCy.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/.claude/skills/hf-dynacell .gemini/skills/hf-dynacell && rm -rf skills-srcUse ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.
Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/
Install the "hf-dynacell" agent skill from https://github.com/mehta-lab/VisCy/tree/main/.claude/skills/hf-dynacell into .gemini/skills/hf-dynacell/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "hf-dynacell", then confirm the skill loads.Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gh skill install mehta-lab/VisCy hf-dynacellInstalls for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).
$ npx skills add mehta-lab/VisCy --skill hf-dynacell -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/mehta-lab/VisCy.git skills-src && mkdir -p .github/skills && cp -r skills-src/.claude/skills/hf-dynacell .github/skills/hf-dynacell && rm -rf skills-srcUse ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.
GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/
Install the "hf-dynacell" agent skill from https://github.com/mehta-lab/VisCy/tree/main/.claude/skills/hf-dynacell into .github/skills/hf-dynacell/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "hf-dynacell", then confirm the skill loads.GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add mehta-lab/VisCy --skill hf-dynacell -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install mehta-lab/VisCy hf-dynacell --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/mehta-lab/VisCy.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/.claude/skills/hf-dynacell .opencode/skills/hf-dynacell && rm -rf skills-srcUse ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.
OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/
Install the "hf-dynacell" agent skill from https://github.com/mehta-lab/VisCy/tree/main/.claude/skills/hf-dynacell into .opencode/skills/hf-dynacell/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "hf-dynacell", then confirm the skill loads.OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
Facts
- Skill name
hf-dynacell- GitHub stars
- 104
- Token cost
- ~1.1k tokens
- SKILL.md length
- 481 words
- Files
- 1
- Skills in repo
- 4
- Repo updated
- First seen
- Licence
- BSD-3-Clause
At a glance
Develop, deploy, and maintain the DynaCell virtual-staining HuggingFace demo hosted at biohub/dynacell (ZeroGPU).
- Works in 4 steps: hf auth login (token with write to the… → Edit hf_space/. → Push: uv run python… → …
- The user asks to push the dynacell space
- SKILL.md covers Hosting — biohub org, Dynacell…, Access model — public demo,…, ZeroGPU configuration… and Code pointers, plus 4 more sections
- Calls hf and uv; needs HF_TOKEN
What it does
Hf Dynacell is an agent skill from mehta-lab/VisCy. Develop, deploy, and maintain the DynaCell virtual-staining HuggingFace demo hosted at biohub/dynacell (ZeroGPU). Covers hosting in the biohub Dynacell resource group, the private-artifacts + HFTOKEN access model, ZeroGPU configuration, and the deploy/smoke-test workflow. Use when the user asks to "push the dynacell space", "update the dynacell demo", "upload checkpoints", "deploy the hf space", mentions biohub/dynacell, or works in applications/dynacell/examples/hfdemo/.
Its SKILL.md is about 1.1k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.
It sits in Testing & QA, covering Model hubs and datasets and QA and bug reports. It works with Hugging Face. The repository describes itself as: computer vision models for single-cell phenotyping. The licence is BSD-3-Clause.
When your agent uses it
- The user asks to push the dynacell space
- Update the dynacell demo
- Upload checkpoints
- Deploy the hf space
Example prompts
- “push the dynacell space”
- “update the dynacell demo”
- “upload checkpoints”
- “/hf-dynacell”
Requirements
- Python 3
Workflow steps
4 steps, taken from the first numbered list in SKILL.md.
- hf auth login (token with write to the Dynacell RG).
- Edit hf_space/.
- Push: uv run python applications/dynacell/examples/hf_demo/upload_hf_space.py.
- Watch: HfApi().get_space_runtime("biohub/dynacell").stage. Build/run logs at
What it can do on your machine
Read from SKILL.md and the folder at commit 4b62365. It shows what the files ask for, not the result of running them.
Tool permissions
Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.
From allowed-tools in the SKILL.md frontmatter.
Runs code
Shell commands in SKILL.md call:
hfuv
From the folder's file list and the shell code blocks in SKILL.md.
Network
No URLs in SKILL.md. Its commands use uv, which can reach the network depending on how they are called.
From URLs in SKILL.md, links to its own repository left out.
Credentials
Names these keys or tokens, usually read from environment variables:
HF_TOKEN
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.
Context cost
Hf Dynacell loads about 1.1k tokens when it runs. Until then it costs about 123 tokens; SKILL.md has 481 words of instructions outside code blocks.
Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.
Safety
Auto-check passed
The automated check found no risky patterns in SKILL.md.
Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.
SKILL.md
The full file from mehta-lab/VisCy at commit 4b62365, republished under its BSD-3-Clause licence (© mehta-lab). 481 words, ~1,138 tokens.
Download SKILL.mdSave it as
.claude/skills/hf-dynacell/SKILL.md (or your agent's skills folder).- name
- hf-dynacell
- description
- Develop, deploy, and maintain the DynaCell virtual-staining HuggingFace demo hosted at biohub/dynacell (ZeroGPU). Covers hosting in the biohub Dynacell resource group, the private-artifacts + HF_TOKEN access model, ZeroGPU configuration, and the deploy/smoke-test workflow. Use when the user asks to "push the dynacell space", "update the dynacell demo", "upload checkpoints", "deploy the hf space", mentions biohub/dynacell, or works in applications/dynacell/examples/hf_demo/.
DynaCell HuggingFace Demo — maintenance
Operational reference for the demo. Content + file layout live in
applications/dynacell/examples/hf_demo/AGENT.md; this skill holds hosting,
access, ZeroGPU, and deploy knowledge.
Hosting — biohub org, Dynacell resource group
Three repos under biohub, all in the Dynacell resource group
(id 6a234bb4507cbbbb04456767):
| Repo | Type | Visibility |
|---|---|---|
biohub/dynacell | Space (Gradio, ZeroGPU) | private (→ public at launch) |
biohub/dynacell-checkpoints | model | private |
biohub/dynacell-demo-data | dataset | private |
RG members who can push: edyoshikun, shalinmehta (admin); alxndrkalinin,
dihan-zheng (write). Migrated from dihan-zheng/* (byte-for-byte mirror).
Moving an existing repo into a resource group needs org-admin. As org-write
members, create repos directly in the group with
create_repo(..., resource_group_id="6a234bb4507cbbbb04456767") — the upload
scripts do this.
Access model — public demo, private artifacts
Decision (2026-06): checkpoints and demo data stay private even when the
Space is public. The Space downloads them server-side via an HF_TOKEN
secret (Space → Settings → Variables and secrets) — a fine-grained,
read-only token scoped to exactly those two repos. Public visitors get only
outputs (plots, PCC); raw artifacts stay gated. If the token is revoked or
loses RG read access, runtime downloads break (rotate via "Replace").
ZeroGPU configuration (validated 2026-06)
- Hardware
zero-a10g(RTX Pro 6000 Blackwell, 48 GB). Enterprise quota 60 GPU-min/day, shared. python_version: "3.12"required in the Space README — the default base image is 3.10, but VisCy needs>=3.12(elseBUILD_ERROR: viscy-data requires ... not in '>=3.12'). ZeroGPU supports Python 3.12.12 / 3.10.13, PyTorch 2.8→latest, Gradio SDK only.requirements.txtincludesspaces;@spaces.GPU(duration=120)decoratesrun_predictionandcompute_trajectory(predict_runner.py).- A subprocess inside
@spaces.GPUinherits the GPU, so the regressiondynacell predictsubprocess works on ZeroGPU.gr.Progressargs work under the decorator. Keepimport torchlazy (no CUDA at import). - Tuning: CELL-Diff at 50–100 ODE steps can approach the 120 s ceiling on a cold
call; raise
durationor use a dynamic-duration callable if timeouts appear.
Show full SKILL.md (192 more words)Show less
Code pointers
CHECKPOINT_REPO (predict_runner.py), _DEMO_REPO (app.py), and the README
models:/datasets: fields all point at biohub/*.
Deploy
hf auth login(token with write to the Dynacell RG).- Edit
hf_space/. - Push:
uv run python applications/dynacell/examples/hf_demo/upload_hf_space.py. - Watch:
HfApi().get_space_runtime("biohub/dynacell").stage. Build/run logs atGET /api/spaces/biohub/dynacell/logs/{build,run}are SSE streams — read withstream=True+ a line cap; a plain GET hangs.
Re-publish checkpoints from HPC: edit the path map in upload_checkpoints.py
and run it where /hpc/projects/... is mounted.
Smoke test
Drive endpoints with gradio_client (Client("biohub/dynacell", token=get_token());
token optional now the Space is public). /load_demo_data renders the data view.
The run endpoints /run_regression and /run_generative read data from gr.State,
which the client resets per call — they can't be driven headlessly, so exercise
them in the live UI. Last validated GPU compute: regression VSCyto3D → Spectral
PCC ~0.95 (~35 s); CELL-Diff trajectory (10 steps) ~74 s.
State
- Space
biohub/dynacellis public (set 2026-06); checkpoints + dataset repos stay private, so theHF_TOKENsecret must remain.
Launch checklist
- Fill citation/DOI TODOs in
cards/*.mdand the uploaded READMEs. - Re-pin
requirements.txtVisCy refs off@dynacell-modelsonce merged/renamed. - Only if the artifact repos are ever made public: the
HF_TOKENsecret can then be removed.
© mehta-lab, BSD-3-Clause. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file
Files
Just SKILL.md in .claude/skills/hf-dynacell of mehta-lab/VisCy.
Open the folder on GitHubat commit 4b62365
Compare with similar skills
Hf Dynacell next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.
| Skill | Stars | Used in | Tokens | Auto-check | Licence | Repo updated |
|---|---|---|---|---|---|---|
| Hf Dynacell this skillmehta-lab/VisCy | 104 | — | ~1.1k | Automated safety check: Pass | BSD-3-Clause | |
| Qwen Mtp GgufR6410418/Jackrong-llm-finetuning-guide | 1.7k | — | ~1.7k | Automated safety check: Pass | MIT | |
| Edge Bringupexeex/edge-cores | 110 | — | ~1.7k | Automated safety check: Notes | Apache-2.0 | |
| Nv Reason CxrNVIDIA/skills | 3.5k | — | ~3.9k | Automated safety check: Notes | Apache-2.0 | |
| Veomni New ModelByteDance-Seed/VeOmni | 2.2k | — | ~2k | Automated safety check: Pass | Apache-2.0 | |
| Setup Benchmark Inputsmlc-ai/pith-train | 355 | — | ~399 | Automated safety check: Pass | Apache-2.0 |
Similar skills
Edge Bringup
exeex/edge-cores
Prepare a macOS or Ubuntu machine for edge-e3 development, diagnose missing Verilator/LLVM/Python dependencies, initialize the public repository, and answer or act on the example prompts in the root…
- Official
Nv Reason Cxr
NVIDIA/skills
Used for command-shape or live NV-Reason-CXR chest X-ray reasoning smoke tests.
Veomni New Model
ByteDance-Seed/VeOmni
A skill your agent uses when adding support for a new model to VeOmni.
Setup Benchmark Inputs
mlc-ai/pith-train
Set up the minimal set of artifacts (tokenized DCLM corpus shard + released HuggingFace checkpoint converted to DCP) required to benchmark, profile, or regression-test a MoE model in PithTrain.
- Official
SageMaker Serving Image Selection
huggingface/skills
Chooses the right serving container and current image URI for deploying a Hugging Face model to a SageMaker endpoint, preferring Hugging Face images over generic ones.
More from mehta-lab/VisCy
Airtable Register
mehta-lab/VisCy
Register zarr positions into the Computational Imaging Database on Airtable, write channelsmetadata/experimentmetadata to zarr .zattrs, or bulk-update Airtable records via MCP.
Prepare Dataset
mehta-lab/VisCy
Prepare datasets for training on VAST storage (NFS - VAST rechunked zarr v3 pipeline).
Airtable Build Collection
mehta-lab/VisCy
Query Airtable MCP to fetch experiment metadata and generate a DynaCLR collection YAML for training
Related
Works with
Categories
Questions about Hf Dynacell
What does Hf Dynacell do?
Develop, deploy, and maintain the DynaCell virtual-staining HuggingFace demo hosted at biohub/dynacell (ZeroGPU). Hf Dynacell is an agent skill from mehta-lab/VisCy. Develop, deploy, and maintain the DynaCell virtual-staining HuggingFace demo hosted at biohub/dynacell (ZeroGPU).
When should I use Hf Dynacell?
Hf Dynacell fits situations like: the user asks to push the dynacell space; update the dynacell demo; upload checkpoints; deploy the hf space.
How do I install Hf Dynacell in Claude Code?
Run `npx skills add mehta-lab/VisCy --skill hf-dynacell -a claude-code`. Or copy the skill folder (.claude/skills/hf-dynacell in mehta-lab/VisCy) into .claude/skills/hf-dynacell in your project. Claude Code loads it when a task matches its description.
How do I install Hf Dynacell in Codex?
Run `npx skills add mehta-lab/VisCy --skill hf-dynacell -a codex`. Or copy the skill folder (.claude/skills/hf-dynacell in mehta-lab/VisCy) into .agents/skills/hf-dynacell in your project. Codex loads it when a task matches its description.
Can I use Hf Dynacell in Cursor, Gemini CLI or GitHub Copilot?
Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add mehta-lab/VisCy --skill hf-dynacell -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/hf-dynacell, .gemini/skills/hf-dynacell, .github/skills/hf-dynacell and .opencode/skills/hf-dynacell in your project.
What does Hf Dynacell need to run?
Going by SKILL.md and its folder, Hf Dynacell needs the command-line tools its instructions call (hf and uv) and credentials named HF_TOKEN. Our summary lists: Python 3.
Does Hf Dynacell access the network?
SKILL.md contains no URLs. Its commands use uv, which can reach the network depending on how they are called. This is read from the text; nothing was executed.
Is Hf Dynacell safe to install?
Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.
What licence does Hf Dynacell use?
Hf Dynacell is published under the BSD-3-Clause licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.
How many tokens does Hf Dynacell use?
About 1.1k tokens (SKILL.md is roughly 4.6k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.
What are the alternatives to Hf Dynacell?
Skills that share tags, products or a category with Hf Dynacell: Qwen Mtp Gguf (R6410418/Jackrong-llm-finetuning-guide, 1.7k stars), Edge Bringup (exeex/edge-cores, 110 stars), Nv Reason Cxr (NVIDIA/skills, 3.5k stars) and Veomni New Model (ByteDance-Seed/VeOmni, 2.2k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
Who maintains Hf Dynacell?
mehta-lab (a GitHub organization) maintains it in mehta-lab/VisCy, which has 104 GitHub stars. The repository holds 4 skills in this directory. The repository was last updated on October 9, 2026.
Source: mehta-lab/VisCy on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.