agentleFS
getappzgetappz/agentflare/.claude/skills/system-architecture-design/SKILL.md

system-architecture-design

Skill3 starsChanged 39 days agoApache-2.0View on GitHub

Use when a change is big enough to need a deliberate structural decision first — new service boundaries, a data model that will be expensive to change later, a technology choice, or anything crossing multiple components — and you want the trade-offs and rationale written down, not just the diagram. Skip for changes confined to one module/file with an obvious approach; skip for the low-level "how do I implement this function" question, which is algorithm-pseudocode-design's job instead.

2 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 4 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Document the decision with the rationale, not just the conclusion.

    line 25
  2. Distinctive:

    Consider non-functional requirements explicitly (performance, security, scalability) — don't let them stay implicit until they cause an …

    line 61
  3. Distinctive:

    Write the ADR even for decisions that feel obvious in the moment.

    line 63
  4. Distinctive:

    Prefer standard diagramming notation (C4, UML) over ad hoc boxes-and- arrows — it's faster for someone else to read.

    line 65

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

System Architecture Design

Produce architecture decisions that explain why, not just what — diagrams and component lists go stale, but a documented trade-off analysis still answers "why did we do it this way" a year later.

When to use

  • Designing service boundaries, a data model, or an API contract that other code will depend on and that's expensive to change once built.
  • Evaluating a technology or pattern choice (microservices vs. monolith, sync vs. event-driven, SQL vs. NoSQL) where the wrong call is costly.
  • Skip for changes scoped to a single module where the approach is already obvious — write the code, don't write an ADR for it.

Key responsibilities

  1. Design scalable, maintainable structure — not the most elaborate one that fits the requirements, the simplest one that does.
  2. Document the decision with the rationale, not just the conclusion.
  3. Produce diagrams that show component interactions and data flow.
  4. Evaluate real trade-offs between options, including "do nothing" and "the boring option."
  5. Account for operational concerns (deployment, monitoring, rollback) — architecture that's elegant to draw but painful to operate is a net loss.

Decision framework

Answer these before committing to a design, not after:

  • What quality attributes actually matter here (throughput, consistency, latency, operability) — and which ones don't, so you're not over-building for them?
  • What are the real constraints and assumptions (team size, existing infrastructure, deadline)?
  • What are the trade-offs of each option under consideration — not just the chosen one?
  • How does this align with (or complicate) existing architecture?
  • What are the risks, and what's the mitigation or rollback plan if the decision turns out wrong?

Deliverables

  1. Architecture Decision Record (ADR) — the decision, the alternatives considered, and why this one won. Short-form is fine; the rationale is the point, not the ceremony.
  2. Component interaction diagram — who talks to whom, and how (sync call, event, shared store).
  3. Data flow diagram — where data originates, transforms, and lands.
  4. Technology evaluation (when applicable) — options compared against the actual requirements, not against feature checklists.

Best practices

  • Consider non-functional requirements explicitly (performance, security, scalability) — don't let them stay implicit until they cause an incident.
  • Write the ADR even for decisions that feel obvious in the moment; the obviousness rarely survives six months.
  • Prefer standard diagramming notation (C4, UML) over ad hoc boxes-and- arrows — it's faster for someone else to read.
  • Design for the extensibility you actually expect, not every hypothetical future requirement — YAGNI applies to architecture too.
  • Think about deployment and monitoring as part of the design, not as a follow-up task.

Handoff

Once the structural decision is made, requirements-specification (if the requirements themselves are still fuzzy) or direct implementation planning via superpowers:writing-plans picks up from here. Use algorithm-pseudocode-design for the lower-level "how does this one function/algorithm work" question — that's a different altitude than a structural decision.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.