From atum-stack-backend
Design scalable distributed systems using structured approaches for load balancing, caching, database scaling, and message queues. Use when the user mentions "system design", "scale this", "high availability", "rate limiter", or "design a URL shortener". Covers common system designs and back-of-the-envelope estimation.
npx claudepluginhub arnwaldn/atum-plugins-collection --plugin atum-stack-backendThis skill uses the workspace's default tool permissions.
A structured approach to designing large-scale distributed systems.
Searches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.
Searches prompts.chat for AI prompt templates by keyword or category, retrieves by ID with variable handling, and improves prompts via AI. Use for discovering or enhancing prompts.
Guides implementation of event-driven hooks in Claude Code plugins using prompt-based validation and bash commands for PreToolUse, Stop, and session events.
A structured approach to designing large-scale distributed systems.
Start with requirements, not solutions. Every system design begins by clarifying what you are building, for whom, and at what scale.
Goal: 10/10. Rate designs 0-10 based on: clear requirements, back-of-the-envelope estimates, appropriate building blocks, scaling/reliability addressed, tradeoffs acknowledged.
| Topic | Reference | Load When |
|---|---|---|
| Estimation | references/estimation.md | Calculating QPS, storage, bandwidth, availability |
| Building Blocks | references/building-blocks.md | DNS, CDN, load balancers, caching, queues, database scaling |
| Common Designs | references/common-designs.md | URL shortener, rate limiter, news feed, chat, search, crawler |
| Mistake | Fix |
|---|---|
| Jump to architecture without requirements | Spend 5-10 min on scope first |
| No estimation | Calculate QPS, storage, bandwidth |
| Single point of failure | Redundancy at every layer |
| Premature sharding | Scale vertically, cache, replicate first |
| Cache without invalidation | Define TTL + explicit invalidation |
| Sync calls everywhere | Queues for non-latency-critical paths |
| No monitoring | Instrument from day one |