mturac

Recsys Pipeline Architect — Data skill for Claude Code

Data community

Composable recommendation pipeline skill for Claude — six-stage Source→Hydrator→Filter→Scorer→Selector→SideEffect framework with Strapi/Go/Python examples.

How to install Recsys Pipeline Architect

This entry records only its repository, not the path inside it, so there is no exact command to give. Open mturac/recsys-pipeline-architect and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Recsys Pipeline Architect does

Composable recommendation pipeline skill for Claude — six-stage Source→Hydrator→Filter→Scorer→Selector→SideEffect framework with Strapi/Go/Python examples.

Alternatives in Data

Portaljs Architect — Recommend a data-portal architecture (storage, compute, catalog, access, hosting, metadata) from your needs, t 2.3k ★
Azure Cosmos DB Python — Cosmos DB development for Python 1.8k ★
Claude Deep Research Skill — Enterprise-grade deep research skill for Claude Code with 8-phase pipeline, source credibility scoring, and au 1k ★

README

recsys-pipeline-architect

A Claude skill for designing composable recommendation, ranking, and feed pipelines — built around the six-stage **Source → Hydrator → Filter → Scorer → Selector → SideEffect** framework popularized by xAI's open-sourced [For You algorithm](https://github.com/xai-org/x-algorithm).

[![License: MIT](https://img.shields.io/badge/license-MIT-blue.svg)](LICENSE) [![Runs on Claude Code](https://img.shields.io/badge/runs%20on-Claude%20Code-orange.svg)](#claude-code) [![Runs on Codex CLI](https://img.shields.io/badge/runs%20on-Codex%20CLI-000000.svg)](#codex-cli) [![Examples](https://img.shields.io/badge/examples-TypeScript%20%7C%20Go%20%7C%20Python-2ea44f.svg)](#examples) [![Pattern: Apache 2.0](https://img.shields.io/badge/pattern%20source-Apache%202.0-lightgrey.svg)](https://github.com/xai-org/x-algorithm) [![skills.sh](https://skills.sh/b/mturac/recsys-pipeline-architect)](https://skills.sh/mturac/recsys-pipeline-architect)

Six-stage recsys pipeline: Source, Hydrator, Filter, Scorer, Selector, SideEffect

Why this exists

Most "recommendation systems" in production aren't exotic ML — they're **pipelines**. You fetch candidates from one or more sources, enrich them with metadata, drop the ineligible, score the rest, sort, pick the top K, then fire off async side effects. The scoring model and the items change. The pipeline shape doesn't.

xAI open-sourced this exact shape in 2024 with their For You algorithm ([Apache 2.0](https://github.com/xai-org/x-algorithm)). This skill turns it into a reusable recipe and applies it well beyond social feeds — Strapi content CMSs, RAG rerankers, task prioritizers, notification triage, search reranking, ad selection: anywhere you need *"the top K items for this (user, context)."*

When you invoke the skill, Claude walks you through eight steps (use case → sources → hydrations → filters → scorers → selector → side effects → scaf