experiment-runner
Standardizes ML experiment tracking and execution.
Install
mkdir -p .claude/skills/experiment-runner && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/13547" && unzip -o skill.zip -d .claude/skills/experiment-runner && rm skill.zipInstalls to .claude/skills/experiment-runner
Activation
This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.
Use when adding, running, or debugging ML experiments, configs, baselines, metrics, or evaluation scripts.Key capabilities
- →Identify experiment entrypoint and config files
- →Add new configs for experiments
- →Add runner hooks for new baselines
- →Implement metric logging for baselines
- →Add README commands for new baselines
- →Perform verification dry-runs for new baselines
How it works
The skill guides the process of adding, running, or debugging ML experiments by identifying entrypoints and configs, preferring new configs over modifying defaults, and ensuring proper setup for new baselines including metric logging and verification dry-runs.
Inputs & outputs
When to use experiment-runner
- →Adding new ML model experiments
- →Configuring training parameters
- →Debugging model evaluation scripts
- →Setting up experiment baselines
About this skill
Workflow:
- Identify the experiment entrypoint.
- Identify config files and output directory.
- Never modify dataset files or checkpoints.
- Prefer adding a new config over changing default configs.
- Keep experiment names explicit.
- For new baselines, add:
- config
- runner hook
- metric logging
- README command
- Verification should use a tiny dry-run if available.
When not to use it
- →When modifying dataset files or checkpoints
- →When changing default configs instead of adding new ones
Limitations
- →Never modifies dataset files or checkpoints
- →Prefers adding new configs over changing default configs
- →Requires explicit experiment names
How it compares
This skill provides a structured workflow for managing ML experiments and baselines, ensuring consistency in configuration, logging, and verification, which is more reliable than ad-hoc experiment management.
Compared to similar skills
experiment-runner side by side with the closest alternatives in the catalog.
| Skill | Installs | Updated | Safety | Difficulty |
|---|---|---|---|---|
| experiment-runner (this skill) | 0 | 1mo | No flags | Intermediate |
| robotics-code-generator | 14 | 7mo | No flags | Advanced |
| modal | 5 | 7mo | Review | Intermediate |
| hugging-face-cli | 3 | 6mo | Review | Intermediate |
Try saying
Example prompts that trigger this skill in your AI assistant.
You might also like
robotics-code-generator
HumaizaNaz
Generates clean, runnable ROS 2, Gazebo, Isaac Sim, and VLA code for humanoid robotics
modal
davila7
Run Python code in the cloud with serverless containers, GPUs, and autoscaling. Use when deploying ML models, running batch processing jobs, scheduling compute-intensive tasks, or serving APIs that require GPU acceleration or dynamic scaling.
hugging-face-cli
patchy631
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.
computer-use-agents
davila7
Build AI agents that interact with computers like humans do - viewing screens, moving cursors, clicking buttons, and typing text. Covers Anthropic's Computer Use, OpenAI's Operator/CUA, and open-source alternatives. Critical focus on sandboxing, security, and handling the unique challenges of vision-based control. Use when: computer use, desktop automation agent, screen control AI, vision-based agent, GUI automation.
machine-learning-ops-ml-pipeline
sickn33
Design and implement a complete ML pipeline for: $ARGUMENTS
ray-train
davila7
Distributed training orchestration across clusters. Scales PyTorch/TensorFlow/HuggingFace from laptop to 1000s of nodes. Built-in hyperparameter tuning with Ray Tune, fault tolerance, elastic scaling. Use when training massive models across multiple machines or running distributed hyperparameter sweeps.