Use in discovery to draw the HIGH-LEVEL architecture in Mermaid — context diagram (C4 L1), containers (C4 L2), and the bounded contexts map (DDD). It reads vision, context-map, design, and assessment, and generates/updates docs/architecture/diagrams.md. It keeps a high level, with no implementation detail. Trigger with /diagramar.
Scanned 9/5/2026
Install to Claude Code
npx -y skills add adriannoes/awesome-agentic-ai --skill diagramar --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Diagramar?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/adriannoes-diagramar)More formats (shields.io, HTML) on the badges page.
---
name: diagramar
description: Use in discovery to draw the HIGH-LEVEL architecture in Mermaid — context diagram (C4 L1), containers (C4 L2), and the bounded contexts map (DDD). It reads vision, context-map, design, and assessment, and generates/updates docs/architecture/diagrams.md. It keeps a high level, with no implementation detail. Trigger with /diagramar.
---
> **Translation note:** Originally authored in Portuguese (pt-BR) by Igor Uehara ([igoruehara/spec-driven](https://github.com/igoruehara/spec-driven), MIT). Translated to English by this hub to keep the repository language consistent. Original content unchanged in meaning; see the upstream repo for the pt-BR source.
# Skill: Diagram architecture (Mermaid, high level)
Generates the architecture diagrams in **Mermaid** (they render on GitHub and in Claude Code) from the
discovery artifacts. Keep it **high level**: actors, systems, containers, and contexts — never classes,
tables, or implementation detail (that is the scope of the feature's `design.md`).
## Inputs (pull only what exists and cite the source)
- `docs/product/vision.md` — actors/personas and the system's purpose.
- `docs/architecture/context-map.md` — bounded contexts and relationships (DDD).
- `specs/*/design.md` — containers, services, and integrations.
- `docs/architecture/assessment.md` — as-is, in brownfield.
- **Gaps** → ask (`AskUserQuestion`): external actors, neighboring systems, the main containers
(UI/services/data/queues), and the critical journey to illustrate.
## Diagrams (high level, aligned with C4 + DDD)
1. **System context (C4 L1):** the system at the center + personas + external systems.
2. **Containers (C4 L2):** apps/services, data, queues, and how they talk.
3. **Bounded contexts map (DDD):** contexts and relationship patterns (ACL, Customer/Supplier,
Conformist, Shared Kernel).
4. *(optional)* **Key flow:** a critical journey in a `sequenceDiagram`.
> Use `flowchart` (renders everywhere) or Mermaid's native C4 diagrams (`C4Context`,
> `C4Container`) if the renderer supports it. Labels in the **ubiquitous language** of `glossary.md`.
## Output
- Write/update `docs/architecture/diagrams.md` (already exists as a placeholder; keep the
frontmatter `alwaysApply: false`). Each diagram in a ` ```mermaid ` block, with a title and 1 line
of context.
- Ensure the back-link in `context-map.md`. **Regenerate when the architecture changes** — an outdated
diagram misleads more than it helps.
## Rules
- **High level only.** Asked for class/table detail? Say it's the scope of `design.md`.
- **Validate the syntax before delivering** — run the deterministic gate (don't trust the eye only):
```
node scripts/validate-mermaid.mjs .
```
It catches empty blocks, missing/unknown diagram types, and unbalanced quotes/delimiters.
Fix the errors before saving; it's the same gate that runs in CI (`esteira.yml`).
- Do not invent components that aren't in the inputs — ask (see knowledge verification in `CLAUDE.md`).
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!