What happens during a model review?

The skill executes an 8-step deep review process that checks your code against per-layer checklists found in the reference documentation to ensure production readiness.

How do I see the current status of my ML models?

You can use the '/haipipe-nn dashboard' command to generate a codebase status table and review the model registry.

Why does the skill require a bootstrap step?

The bootstrap step ensures Claude reads the overview.md file first, providing the necessary architecture map and YAML templates required for accurate execution.

Can this skill help with specific model types like bandits or diffusion?

Yes, haipipe-nn is specifically designed to handle mlpredictor, tsforecast, tefm, tediffusion, and bandit models.

What are the four layers supported by haipipe-nn?

The skill manages four distinct layers: Algorithm (L1), Tuner (L2), Instance (L3), and ModelSet/Pipeline (L4), allowing for granular control over the entire ML stack.

Haipipe NN Pipeline Manager

Name: Haipipe NN Pipeline Manager
Author: jluo41

byjluo41

0•

データサイエンスとML

Manages and orchestrates end-to-end neural network pipelines across algorithmic, tuning, and deployment layers.

haipipe-nn is a comprehensive Claude Code skill designed for machine learning engineers to manage the full lifecycle of neural network models within the hainn codebase. It provides a structured four-layer framework—Algorithm, Tuner, Instance, and ModelSet—to handle everything from codebase status monitoring and deep architectural reviews to automated code generation and rigorous testing protocols. Whether working on time-series forecasting, bandits, or diffusion models, this skill ensures consistency through standardized YAML templates, registry checks, and a mandatory bootstrap protocol that aligns Claude with your specific architecture.

主な機能

01Four-layer architectural oversight spanning Algorithm, Tuner, Instance, and ModelSet layers.

020 GitHub stars

03Deep 8-step model review protocol to ensure architectural integrity and contract compliance.

04Integrated 7-step test design and execution protocol for specialized NN pipeline debugging.

05Automated code generation guide for rapid development of new ML models and components.

06Unified dashboard for real-time codebase status scans and model registry tracking.

ユースケース

01Monitoring model status and registry consistency across large-scale machine learning repositories.

02Generating standardized boilerplate and implementation code for new neural network algorithms or tuners.

03Conducting comprehensive architectural reviews of complex ML pipelines to ensure adherence to framework contracts.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add jluo41/tools haipipe-nn

For use in Claude.ai and ChatGPT

Download Skill