init-physics-intern

Scaffold a PhysicsIntern research workspace in the current folder.

npx skills add https://github.com/huggingface/physics-intern-claude-plugin --skill init-physics-intern

The PhysicsIntern initializer script is located at:

!echo "${CLAUDE_PLUGIN_ROOT}/scripts/plugin-init.sh"

This is a purely mechanical bootstrap. Do exactly the following and nothing more — do not read, copy, render, or create any files yourself, and do not narrate steps:

  1. First, print this line verbatim to the user:

    ⏳ Initializing the PhysicsIntern workspace. This needs one permission approval — the script writes .claude/ and runs git init, which the sandbox blocks by default. Please approve it when prompted.

  2. Then run that one script, with no arguments, using the Bash tool. The script renders the workspace, scaffolds the directories, and makes the first git commit — it does the entire job in one command. It writes to .claude/ and runs git init, which the sandbox blocks, so Claude Code will ask to run it outside the sandbox (or for permission) — that is expected; approve it. Do not fall back to doing the work by hand if it is denied — just re-run the same script with permission.

  3. Read the script's final RESULT: line, then:

    • If it contains RESULT: initialized, print the following message verbatim and then stop — add nothing else, no file lists, no extra suggestions:

      ✅ PhysicsIntern workspace initialized. Two steps to begin:

      1. Edit problem.md — fill in the problem setup and main question.
      2. Exit and restart Claude Code in this folder, then run /survey.
    • If it contains RESULT: already-initialized, tell the user this folder already contains a PhysicsIntern workspace and was left untouched, and stop.

    • If there is no RESULT: line, show the script's output to the user and stop.

More skills from huggingface

sync-models
huggingface
Sync chat-ui's model config with the HuggingFace router — add descriptions for new models, flag reasoning-capable ones, enable artifacts for models with 32B+…
custom-blocks
huggingface
Use when the user has written (or wants to write) a `ModularPipelineBlocks` subclass in a local Python file and needs to package it into a Hub-uploadable…
self-review
huggingface
Use before opening a PR, or whenever asked to self-review a diffusers contribution. Applies the same rubric as the `@claude` CI (checks the diff against…
hf-cloud-sagemaker-production-defaults
huggingface
Create a SageMaker endpoint (real-time or async) with autoscaling, CloudWatch alarms, and tagging enabled by default. Use this skill whenever about to create a…
hf-cloud-serving-image-selection
huggingface
Pick the right serving container for a SageMaker model deployment and find its current image URI. Use this skill whenever about to deploy a model to a…
Hugging Face Cli
huggingface
Execute Hugging Face Hub operations using the `hf` CLI. Use when the user needs to download models/datasets/spaces, upload files to Hub repositories, create repos, manage local cache, or run compute jobs on HF infrastructure. Covers authentication, file transfers, repository creation, cache operations, and cloud compute.
Hugging Face Datasets
huggingface
Create and manage datasets on Hugging Face Hub. Supports initializing repos, defining configs/system prompts, streaming row updates, and SQL-based dataset querying/transformation. Designed to work alongside HF MCP server for comprehensive dataset workflows.
Hugging Face Evaluation
huggingface
Add and manage evaluation results in Hugging Face model cards. Supports extracting eval tables from README content, importing scores from Artificial Analysis API, and running custom model evaluations with vLLM/lighteval. Works with the model-index metadata format.