enforcing-seed-hygiene
CommunityKeep ML runs reproducible, every time.
Authorrocklambros
Version1.0.0
Installs0
System Documentation
What problem does it solve?
This Skill prevents flaky, non-reproducible results in notebooks and ML pipelines by making randomness explicit and consistent across runs, libraries, and platforms.
Core Features & Use Cases
- Seeds the full stack in one first-cell block, including Python random, NumPy, PyTorch, JAX, TensorFlow, and R.
- Handles cross-platform sampler reproducibility with CPU and thread pinning when fixed-seed behavior is not enough.
- Adds repo-level guardrails with a pre-commit check that catches new files missing seed setup.
- Use it when starting a new notebook, debugging run-to-run drift, preparing reproducible analysis, or standardizing ML project templates.
Quick Start
Use the enforcing-seed-hygiene skill to generate a single first-cell reproducibility setup for my new PyTorch notebook.
Dependency Matrix
Required Modules
None requiredComponents
references
💻 Claude Code Installation
Recommended: Let Claude install automatically. Simply copy and paste the text below to Claude Code.
Please help me install this Skill: Name: enforcing-seed-hygiene Download link: https://github.com/rocklambros/rcs/archive/main.zip#enforcing-seed-hygiene Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
Agent Skills Search Helper
Install a tiny helper to your Agent, search and equip skill from 536,000+ vetted skills library on demand.