Architecture

Skill KhaiTrang1995/agentic-awesome-kits/packages/agentic-awesome-kits/payload/kits/software/drawio-kit/.claude/skills/architecture

A universal collection of reusable Loop Engineering kits for AI coding agents—not just prompts. Each kit combines Skills, Rules, Templates, workflows, and validation loops to help agents autonomously plan, execute, verify, fix, and improve engineering tasks.

Install

npx -y skills add KhaiTrang1995/agentic-awesome-kits --skill architecture

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

3 things to look at

22 days oldThe repository was created 22 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
7 stars7 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Generate system architecture .drawio diagrams — client, service, database, external systems with standard icons and clear connections.

SKILL.md

2.9 KB, as published. Nobody here has run it

/architecture — Generate Architecture Diagram (draw.io)

When to use

Visualize system architecture (microservice, layered architecture, network topology) for design docs, ADRs, or stakeholder presentations.

Input examples

/architecture TOPO system: Angular SPA → API Gateway → 3 microservices (Container, Booking, Auth) → SQL Server
/architecture --cloud azure deploy architecture: App Service, Azure SQL, Blob Storage, Front Door
/architecture integration flow with Customs system via REST API

Workflow

Phase 1 — Identify components & trust boundaries

List components by group:

Client: web app, mobile app, external user (actor)
Edge: load balancer, API gateway, CDN
Service: backend service/microservice (process box)
Data: database, cache, message queue (cylinder/tape)
External: third-party system (cloud shape)

Identify trust boundaries (internet ↔ DMZ ↔ internal) — draw with container boxes dashed=1;fillColor=none;.

Phase 2 — Layout (L3 preview)

Print a block diagram as table/ASCII before generating XML — confirm which groups sit inside which boundaries and the data-flow direction.

Phase 3 — Generate XML

Use _templates/architecture.drawio as the frame:

Default to core shapes (ellipse+shape=cloud, shape=cylinder3, shape=umlActor, rounded=1) — opens on any machine.
If --cloud aws|azure|gcp: add matching stencils (shape=mxgraph.aws4.*, shape=mxgraph.azure.*) and note clearly in the output that the user must enable that shape library in draw.io.
Group by trust boundary with surrounding container boxes.
Primary data flow left→right or top→bottom, orthogonal arrows.

Phase 4 — Approval L1 → Write

Mandatory rules

Every external system/3rd-party MUST be visually distinct with gray (fillColor=#f5f5f5;strokeColor=#666666;).
Databases always use shape=cylinder3 — NEVER use a rounded rectangle for storage.
If there are > 10 components, group by layer/boundary with container boxes for readability instead of leaving all at the same level.
Small legend in the lower corner if using > 3 color types, explaining each color.

Output

docs/diagrams/{feature}/architecture-{slug}.drawio

References

@.claude/rules/drawio-conventions.md
@.claude/rules/naming-conventions.md
@.claude/rules/approval-gate.md
@_templates/architecture.drawio
Brain: DRAWIO-BRAIN.md