ai4science-studio: Command for Claude Code

.claude/commands/add-recipe.md

add-recipe is a command for Claude Code from AMDResearch/ai4science-studio. It costs 0 tokens per session (876 once invoked), scanned A, original, MIT.

A workflow for creating or updating instructions and example scripts for using, fine-tuning, or evaluating an existing AI model in AI4Science Studio. Fine-tuning means adapting a model to a particular task, while evaluation means testing how well it performs.

In plain words
What is it for?
Use it to add inference, fine-tuning, or evaluation recipes, example shell or Python scripts, container startup steps, and preflight checks for GPU access and imports.
Why use it?
It gives model users a consistent recipe to follow and helps keep setup, configuration, and basic checks in one place.

Command for Claude Code

Written for Claude Code: $ARGUMENTS substitution.

This is AMDResearch/ai4science-studio's own configuration. It tells Claude Code how to work on ai4science-studio itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything ai4science-studio configures →

Reuse

Borrowing it

Nothing to install: this file belongs to AMDResearch/ai4science-studio. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/AMDResearch/ai4science-studio/main/.claude/commands/add-recipe.md
Clone the repo
git clone --depth 1 https://github.com/AMDResearch/ai4science-studio

Made for: Claude Code.

Wrote 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.

agentmods badge for add-recipe

README.md
[![agentmods](https://agentmods.dev/badge/commands/amdresearch/ai4science-studio/add-recipe.svg)](https://agentmods.dev/commands/amdresearch/ai4science-studio/add-recipe)
Your own site
<a href="https://agentmods.dev/commands/amdresearch/ai4science-studio/add-recipe"><img src="https://agentmods.dev/badge/commands/amdresearch/ai4science-studio/add-recipe.svg" alt="Measured on agentmods" height="20"></a>
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 876 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00000 $0.00876
Opus 5 $0.00000 $0.00438
Sonnet 5 $0.00000 $0.00175
Haiku 4.5 $0.00000 $0.00088

Measured 6d ago against content hash 720fa4e86fdb, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

add-recipe 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 6d 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.

.claude/commands/add-recipe.md · 58 lines

How it starts

The opening of the file, as written. The whole thing — 58 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Add or update a recipe for an existing model

Create or improve a recipe (inference, fine-tune, eval, etc.) for a model already in AI4Science Studio.

Steps

  1. Locate the model folder<domain>/models/<slug>/.

  2. Choose or create a recipe subfolder under recipes/:

    • recipes/inference/
    • recipes/finetune/
    • recipes/eval/
    • other task-specific names as needed
    • Each subfolder needs a README.md with a callout box at the top linking to examples/.
  3. Write or update example scripts under <slug>/examples/:

    Script Purpose
    docker_run.sh Launch the container; auto-detect AMD Container Toolkit (docker info | grep -qi amd) vs device passthrough (/dev/kfd + all /dev/dri/renderD* nodes); check for existing container and exit with attach hint rather than error; auto-clone upstream repo if absent.
    run_<task>.sh / run_<task>.py Run the model; all key params overridable via env vars with sensible defaults; print config summary before running; print error + fix instructions when required inputs are missing, then exit 1.
    preflight_<slug>.py Smoke-test: verify GPU access, imports, and a tiny forward pass.
    sbatch_<task>_amd.sh SLURM batch script; use --rocm (not --nv) for AMD/Apptainer GPU passthrough. Named _amd.sh (not _mi300x.sh) — same script works on MI250X, MI300X, and MI350X with rocm7.2.x.
    build_overlay_amd.sh (HPC models with heavy pip deps) Builds a persistent Apptainer ext3 overlay. Install to NFS staging dir, strip torch/nvidia/triton, copy to overlay. Expose ROCM_WHL_TAG as the ROCm version knob.
    • All scripts must be chmod +x.
    • Use a tl_config.toml.template alongside run scripts for config-file-driven models.
  4. Recipe README content:

    • Add a callout box at the top pointing to examples/.
    • Pin Python/framework versions.
    • Include ROCm/CUDA notes only for validated results; state what was tested and the outcome (e.g. speedup %).
    • Reuse upstream scripts with minimal wrappers; avoid reimplementing full stacks.

Read the full file on GitHub · 58 lines

Changes

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.

  1. 6d ago First seen · 58 lines · 0 tokens per session scan A 720fa4e86fdb

Subscribe to this mod's changes

add-recipe is a command published in the GitHub repository AMDResearch/ai4science-studio (4 stars, last pushed 1mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 876 tokens. 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 2026-08-31.