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 skills add NVIDIA-TAO/tao-skill-bank --skill tao-finetune-anomalygennextgit clone --depth 1 https://github.com/NVIDIA-TAO/tao-skill-bankWrote 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/skills/nvidia-tao/tao-skill-bank/tao-finetune-anomalygennext)<a href="https://agentmods.dev/skills/nvidia-tao/tao-skill-bank/tao-finetune-anomalygennext"><img src="https://agentmods.dev/badge/skills/nvidia-tao/tao-skill-bank/tao-finetune-anomalygennext/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/nvidia-tao/tao-skill-bank/tao-finetune-anomalygennext"><img src="https://agentmods.dev/badge/skills/nvidia-tao/tao-skill-bank/tao-finetune-anomalygennext.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00058 | $0.00858 |
| Opus 5 | $0.00029 | $0.00429 |
| Sonnet 5 | $0.00012 | $0.00172 |
| Haiku 4.5 | $0.00006 | $0.00086 |
Grade A, and why
tao-finetune-anomalygennext 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 2d 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 — 93 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Fine-tune AnomalyGenNext
This leaf prepares a user-owned dataset and optional recipe for task-specific
AnomalyGenNext LoRA training. The pinned image is declared in
references/skill_info.yaml; do not replace it with the older 1.0 release.
Read references/input-contract.md when validating a dataset or adapting a
recipe, and references/container-runtime.md before submission.
Inputs
DATASET/
defect_spec.jsonl
TEXTURE/
clean_image/*
anomaly_image/DEFECT/*
mask/DEFECT/*
VALIDATION/testcase.jsonl
Also supply the Cosmos3-Nano base checkpoint directory, Wan2.2_VAE.pth, and
a local facebook/dinov2-large checkpoint directory. The validation JSONL must
contain image_filename, mask_filename, and anomaly_type; each trained
TEXTURE+DEFECT needs at least three rows. A separately stored defect spec is
accepted with --defect-spec.
An optional user recipe.yaml is a template. Custom training settings remain,
while dataset, checkpoint, validation, type order, and iteration-zero validation
are replaced by validated identities. Without a template, the packaged recipe
uses the established 5000-step defaults.
Prepare
After the common launch review, run the prepare_recipe action:
scripts/prepare_finetune_recipe.py \
--dataset-root /data/my_dataset \
--validation-testcase /data/validation/testcase.jsonl \
--base-checkpoint /models/Cosmos3-Nano \
--vae-path /models/Wan2.2_VAE.pth \
--nn-backbone /models/facebook/dinov2-large \
--dataset-name my_dataset \
--recipe-template /data/recipe.yaml \
--output /results/canonical_recipe.yaml
The action freezes absolute validation paths, validates anomaly images and
masks, checks type agreement with defect_spec.jsonl, refuses output reuse,
and emits a recipe plus metadata. validation_iter must be a multiple of
save_iter; max_iter must reach a post-baseline validation.
Train and accept
Invoke tao-launch-workflow, review the platform, image, mounts, GPU shape,
runtime, and exact recipe, then submit the train action. Bind the selected
DINOv2 directory read-only at the fixed container path declared in
skill_info.yaml; the public image does not bundle it. An optional Hugging Face
cache supplies the Qwen tokenizer for offline execution.
What ships with it
11 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- assets/default_recipe.yaml 593 B
- evals/evals.json 806 B
- references/container-runtime.md 1.3 KB
- references/input-contract.md 1.9 KB
- references/nn-score-contract.md 837 B
- references/skill_info.yaml 2.5 KB
- scripts/finetune_anomalygennext.sh 3.1 KB runs code
- scripts/prepare_finetune_recipe.py 10.0 KB runs code
- scripts/tests/test_nn_improvement.py 2.6 KB runs code
- scripts/tests/test_prepare_finetune_recipe.py 3.3 KB runs code
- scripts/validate_nn_improvement.py 6.4 KB runs code
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.
- 2d ago First seen · 93 lines · 58 tokens per session scan A a90785dc1529
tao-finetune-anomalygennext is a skill published in the GitHub repository NVIDIA-TAO/tao-skill-bank (88 stars, last pushed today), licensed Apache-2.0. It adds 58 tokens to every session and 858 once invoked, about $0.0003 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 2026-09-10.
Other skills, from other repositories
huawei-cloud-cloudrobo-train
Manage CloudRobo model training tasks and simulation reinforcement learning (SimRL) tasks — create pretrain (TRAINFROMSCRATCH) and finetune (MODELTUNING) tasks with FFT/SFT/LORA/QLORA/DEEPSPEED methods; manage the full task lifecycle (create/read/update/delete/stop/restart/resume/draft); save and resubmit draft…
tinker-research
Pi overlay for Tinker post-training. Use when the user wants to plan SFT/RL/DPO/distillation, pick a model, or run experiments from Pi. Canonical methodology lives in Tinker Cookbook — do not reinvent it here.
huggingface-transformers
Use Hugging Face Transformers pipelines, tokenizers, and AutoModel interfaces for inference and fine-tuning workflows.
Training
Guides fine-tuning and post-training work with explicit data, objective, hardware, and rollback assumptions.
Data Validation
Data quality checks to run before modeling — distribution summaries, leakage detection, class balance.
ml-research-methodology
Use at the START of ANY machine-learning / deep-learning / AI modeling task - building, training, fine-tuning, or choosing a model for image classification, object/face/vehicle detection, segmentation, medical imaging (tumor/cancer/MRI/X-ray/mammogram), text/NLP/LLM, tabular prediction (churn, price, risk), or…