Engineering ArchitectureDesignSystems

Senior Architect

System architecture design, dependency analysis, and technical decisions.

$skillfish add ~skills/senior-architect

Real Examples Get Started Models & Context

Real Conversations

See how the Senior Architect skill responds to actual requests. These are representative examples of what your AI agent will produce.

You

Redesigning our Rails monolith handling ~5,000 orders/day. Checkout takes 8-12 seconds. Looking at event-driven microservices. Where do I start?

Agent

Architecture: Strangler Fig + Async Checkout Core

Don't rip out the monolith — strangle it. Extract the slowest path (checkout) first.

Why checkout is 8-12s (almost certainly): Synchronous chain: inventory check → Stripe charge → send email → send SMS. Each waits for the previous. Event-driven breaks the chain after payment.

Target architecture:

Checkout API (Go/Node) → Kafka: orders.created
                           ├→ Inventory Service  (reserve stock)
                           ├→ Fulfillment Service (pick/pack)
                           └→ Notification Service (email + SMS)

Rails monolith: catalog, user mgmt, admin — untouched in Phase 1

Phase 1 (Month 1-2) — fix latency:

Thin checkout API: validates cart, charges Stripe, publishes order.created event
Email/SMS move to async Kafka consumers
Checkout response time: 8-12s → <800ms

Phase 2 (Month 3-4) — inventory + fulfillment:

Saga pattern for distributed transaction: reserve → charge → fulfill
Compensating transactions: release reservation if charge fails

Kafka topic sizing for 5k orders/day:

Topic	Partitions	Retention
orders.created	6	7 days
orders.failed	3	30 days

Rule: Add a new service only when a bounded context needs different scaling OR separate team ownership. Two services in Month 1 is enough.

You

15-engineer team, 4 frontend apps, 3 shared component libraries, 2 Python services, 1 mobile app. Polyrepo now and dependency management is chaos. Should we switch?

Agent

Decision: Monorepo with Turborepo

Version drift in shared libraries and coordinated cross-repo PRs are the canonical monorepo problems. At 15 engineers you'll get full benefit without enterprise-scale pain.

Why Turborepo over Nx:

Simpler config for JS/Python mixed stack (Nx's Python support is awkward)
Remote caching via Vercel (free under 10GB/month)
Incremental adoption — you don't rewrite CI in one shot

Proposed workspace structure:

/apps
  web-marketing/   # Next.js
  web-app/         # Next.js
  web-admin/       # React
  mobile/          # React Native
/packages
  ui/              # shared component library
  design-tokens/   # Tailwind config, color palette
  api-client/      # generated OpenAPI types + fetch wrappers
/services
  api/             # Python FastAPI (isolated venv)
  workers/         # Python Celery (isolated venv)

Migration — don't big-bang it:

Week 1: scaffold monorepo, import one frontend app, configure Turborepo cache
Week 2-3: migrate the shared UI library — highest-value step, unblocks everything else
Week 4+: migrate apps one at a time

Risks to plan for:

Git history: use git filter-repo to preserve it — do not squash on import
CI time: configure Turborepo remote cache before migrating or CI will slow down temporarily
Python: keep independent virtualenvs per service — don't force a shared Python toolchain

Get Started

How to install and use this skill in your preferred environment.

Skills are designed for AI coding agents (Claude Code, Cursor, Windsurf) and IDE-based workflows where the agent can read files, run scripts, and act on your codebase. Web-based AI can use the knowledge and frameworks, but won't have tool access.

Models & Context

Which AI models and context windows work best with this skill.

Recommended Models

Best

Claude Opus 4 Claude Sonnet 4 GPT-4.1 Gemini 2.5 Pro Grok 3 Kimi K2

Good

Claude Haiku 4.5 GPT-4.1 mini Gemini 2.5 Flash Grok 3 mini

Larger models produce more detailed, production-ready outputs.

Context Window

This skill's SKILL.md is typically 3–10 KB — fits in any modern context window.

8K Skill only

32K+ Skill + conversation

100K+ Skill + references + codebase

All current frontier models (Claude, GPT, Gemini) support 100K+ context. Use the full window for complex multi-service work.

Pro tips for best results

1

Be specific

Include numbers — users, budget, RPS — so the skill can size the architecture.

2

Share constraints

Compliance needs, team size, and existing stack all improve the output.

3

Iterate

Start with a high-level design, then ask follow-ups for IaC, cost analysis, or security review.

4

Combine skills

Pair with companion skills below for end-to-end coverage.

Ready to try Senior Architect?

Install the skill and start getting expert-level guidance in your workflow — any agent, any IDE.

$skillfish add ~skills/senior-architect

← Browse all 169 skills