Can I review existing pipeline code for errors?

Yes, the 'review' command performs a structural code review by detecting the stage type and applying specific MUST-DO and MUST-NOT checklists.

What are the four stages of the Haipipe pipeline?

The pipeline follows a four-stage process: 1-Source (ingestion), 2-Record (temporal alignment), 3-Case (feature naming and triggers), and 4-AIData (transformations and dataset splits).

What is the difference between 'design-chef' and 'design-kitchen'?

'design-chef' is used to create specific processing functions (Fns) for data, while 'design-kitchen' is used to modify the underlying pipeline infrastructure itself.

How does the 'cook' command ensure safety?

The 'cook' command follows a strict protocol where it displays the complete YAML configuration and requires explicit user confirmation before any pipeline execution begins.

Haipipe Data Pipeline Manager

Name: Haipipe Data Pipeline Manager
Author: jluo41

byjluo41

0•

데이터 과학 및 ML

Manages end-to-end data pipelines across four stages from raw source ingestion to AI-ready datasets.

Haipipe Data is a comprehensive Claude Code skill designed to orchestrate complex data engineering workflows through a structured four-stage architecture: Source, Record, Case, and AIData. It provides developers with a unified interface to inspect workspace assets via a dashboard, design custom processing functions (such as SourceFn or RecordFn) using a guided 'chef' metaphor, and execute pipeline runs with strict configuration validation. By enforcing a step-by-step protocol and providing structural code reviews, it ensures that data flows correctly from raw files to machine learning-ready formats.

주요 기능

01Structural code review and concept explanation for pipeline architecture

02Interactive 'Design-Chef' mode for building custom processing functions

030 GitHub stars

04Unified 4-stage pipeline management (Source, Record, Case, AIData)

05Strict 'Cook' protocol requiring manual YAML configuration sign-off before execution

06Real-time workspace dashboard for monitoring asset status and manifest checks

사용 사례

01Building and testing data transformation functions for AI training sets

02Monitoring and managing the lifecycle of large-scale data processing jobs

03Auditing complex data pipelines for schema compatibility and structural correctness

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add jluo41/tools haipipe-data

For use in Claude.ai and ChatGPT

Download Skill