tao-validate-recipe-transfer agent skill by NVIDIA

At a glance

Before you install

Read from the skill's files at the directory's October 19, 2026 snapshot: what ships with it, what its code reads, and the terms it is published under.

• Bundled scripts: 4 script files (4 Python). Your agent may run them, so read them first.
• Tool permissions requested: Read Bash Write (the allowed-tools field of SKILL.md).
• License: Apache-2.0, an open-source or open-content license.

Install tao-validate-recipe-transfer

The skills CLI works with Claude Code, Codex, Cursor and other agents; the other routes copy the folder by hand.

npx skills add https://github.com/NVIDIA/skills/tree/main/skills/tao-validate-recipe-transfer

tmp="$(mktemp -d)"
git clone --depth 1 https://github.com/NVIDIA/skills.git "$tmp/repo"
mkdir -p .claude/skills
cp -R "$tmp/repo/"skills/tao-validate-recipe-transfer .claude/skills/
rm -rf "$tmp"

python3 ~/.codex/skills/.system/skill-installer/scripts/install-skill-from-github.py \
    --repo NVIDIA/skills \
    --path skills/tao-validate-recipe-transfer

NVIDIA/skills is third-party code. Read SKILL.md and any bundled scripts before you install it.

How the author describes it

Port a published computer vision paper's official code and training recipe onto a customer's own dataset, or diagnose why such a transfer produced bad numbers. Use this whenever someone wants to reproduce a CV paper, run a paper's repo on their own images, fine-tune a published detection/segmentation/classification/keypoint model on customer data, adapt a training recipe to a new dataset, or figure out why a fine-tuned vision model scores well on validation but fails in production. Also use for post-mortems on any failed or disappointing CV training run, and whenever a user mentions mAP that looks too good, a model that "worked in training but not in deployment", or transferring hyperparameters from a paper to their own data. Trigger even if the user only says "train a model on my dataset" and a published architecture or repo is involved.

From the description field of SKILL.md. Agents read this line to decide when to load the skill.

Compatibility, from SKILL.md

Sections in SKILL.md

• The core idea
• Mode selection
• Phase 0 — Feasibility gate
• Phase A — Port
• Phase G — Domain gap report
• Phase T — Transfer the recipe
• The trap that catches almost everyone
• Preprocessing is a single artifact
• Phase R — Run
• Phase E — Evaluate and report
• Reference files
• Scripts
• Honest framing

Files in the skill folder

• BENCHMARK.md
• SKILL.md
• assets/status.example.json
• config/skillspector-baseline.yaml
• evals/evals.json
• references/failure-atlas.md
• references/port-gate.md
• references/postmortem.md
• references/recipe-fields.md
• references/recipe_spec.yaml
• references/stacks.md
• references/verification-ladder.md
• scripts/domain_gap_report.py
• scripts/preprocess_parity.py
• scripts/render_status.py
• scripts/split_leakage_check.py
• skill-card.md
• skill.oms.sig

Suede Skill Check: 88/100 · B

tao-validate-recipe-transfer passes the required frontmatter check and all seven graded checks, which cover the SKILL.md format, a declared license and a commit in the last year. Those checks earn 56 of the score’s points; the rest come from description quality, instruction depth, bundled files, maintenance and curated-list citations. How the score works.

Where this skill is cited

NVIDIA/skills appears in 1 curated list: VoltAgent/awesome-agent-skills.

More AI agent skills

All AI agent and LLM engineering skills

More from NVIDIA/skills

All 396 skills in NVIDIA/skills