mermaid-studio

---
name: mermaid-studio
description: Expert Mermaid diagram creation, validation, and rendering with dual-engine output (SVG/PNG/ASCII). Supports all 20+ diagram types including C4 architecture, AWS architecture-beta with service icons, flowcharts, sequence, ERD, state, class, mindmap, timeline, git graph, sankey, and more. Features code-to-diagram analysis, batch rendering, 15+ themes, and syntax validation. Use when users ask to create diagrams, visualize architecture, render mermaid files, generate ASCII diagrams, document system flows, model databases, draw AWS infrastructure, analyze code structure, or anything involving "mermaid", "diagram", "flowchart", "architecture diagram", "sequence diagram", "ERD", "C4", "ASCII diagram". Do NOT use for non-Mermaid image generation, data plotting with chart libraries, or general documentation writing.
license: CC-BY-4.0
metadata:
  author: Felipe Rodrigues - github.com/felipfr
  version: 1.0.1
---

# Mermaid Studio

Expert-level Mermaid diagram creation, validation, and multi-format rendering. Creates diagrams from descriptions or code analysis, validates syntax, and renders to SVG, PNG, or ASCII with professional theming.

## Golden Rules — Elegant Diagrams by Default

Every diagram MUST follow these principles. They are not optional — they define the difference between a mediocre diagram and a gold-standard one.

### Rule 1: Always Use an Init Directive for Professional Styling

**NEVER** create a diagram without an `%%{init}` directive or frontmatter config. The default Mermaid theme produces harsh black lines and generic colors. Always apply a curated palette.

**For general diagrams (flowchart, sequence, state, class, ERD):**

```
%%{init: {'theme': 'base', 'themeVariables': {
  'primaryColor': '#4f46e5', 'primaryTextColor': '#ffffff',
  'primaryBorderColor': '#3730a3', 'lineColor': '#94a3b8',
  'secondaryColor': '#10b981', 'tertiaryColor': '#f59e0b',
  'background': '#ffffff', 'mainBkg': '#f8fafc',
  'nodeBorder': '#cbd5e1', 'clusterBkg': '#f1f5f9',
  'clusterBorder': '#e2e8f0', 'titleColor': '#1e293b',
  'edgeLabelBackground': '#ffffff', 'textColor': '#334155'
}}}%%
```

**⚠️ Font Warning:** Do NOT set `fontFamily` in theme variables. The Mermaid default font (`trebuchet ms, verdana, arial, sans-serif`) works everywhere. Setting `system-ui`, `Segoe UI`, or `-apple-system` will render as Times New Roman in headless Chromium (used by `mmdc`).

**For C4 diagrams — see the dedicated C4 styling section below.**

**For architecture-beta diagrams — see the dedicated AWS/Architecture section below.**

### Rule 2: Soft Lines, Never Harsh Black

The single biggest visual improvement is using `lineColor: '#94a3b8'` (slate-400) instead of the default black. This creates a modern, breathable diagram. For dark themes, use `lineColor: '#64748b'` (slate-500).

### Rule 3: Limit Density — Breathe

- Maximum 15 nodes per diagram (not 20 — fewer is more elegant)
- Use `subgraph` or boundaries to create whitespace and visual grouping
- Prefer LR (left-right) for process flows — it reads more naturally
- Use invisible links (`A ~~~ B`) to add spacing when the layout is cramped

### Rule 4: Meaningful Labels and Consistent Style

- Node IDs: camelCase (`orderService`, not `s1` or `os`)
- Labels: short, clear natural language (`[Order Service]`)
- Arrows: action verbs with protocol info (`"Sends order via gRPC"`)
- Descriptions: one-line, role-focused (`"Handles order lifecycle"`)

### Rule 5: Color Harmony Over Color Variety

Use max 3-4 colors per diagram. Map colors to meaning:

- **Blue tones** (#4f46e5, #3b82f6) → primary systems, internal services
- **Green tones** (#10b981, #059669) → success states, data stores
- **Amber tones** (#f59e0b, #d97706) → external systems, warnings
- **Slate tones** (#64748b, #94a3b8) → lines, borders, secondary elements
- **Red tones** (#ef4444) → errors ONLY, never as decoration

## Modes of Operation

This skill operates in three modes based on user intent:

| Mode       | Trigger                                           | What happens               |
| ---------- | ------------------------------------------------- | -------------------------- |
| **Create** | "draw a diagram of...", "visualize my..."         | Generates .mmd code only   |
| **Render** | "render this mermaid", "convert to SVG/PNG/ASCII" | Renders existing .mmd      |
| **Full**   | "create and render...", ambiguous requests        | Create → Validate → Render |

Default to **Full** mode when intent is unclear.

## Step 1: Understand the Request

Before writing any Mermaid code, determine:

1. **What to diagram** — system, flow, schema, architecture, code structure?
2. **Which diagram type** — use the Decision Matrix below
3. **Output format** — code only, SVG, PNG, or ASCII?
4. **Theme preference** — ask only if rendering; default to `base` theme with curated palette

### Diagram Type Decision Matrix

| User describes...                              | Diagram Type  | Syntax keyword       |
| ---------------------------------------------- | ------------- | -------------------- |
| Process, algorithm, decision tree, workflow    | Flowchart     | `flowchart TD/LR`    |
| API calls, message passing, request/response   | Sequence      | `sequenceDiagram`    |
| Database schema, table relationships           | ERD           | `erDiagram`          |
| OOP classes, domain model, interfaces          | Class         | `classDiagram`       |
| State machine, lifecycle, transitions          | State         | `stateDiagram-v2`    |
| High-level system overview (C4 Level 1)        | C4 Context    | `C4Context`          |
| Applications, databases, services (C4 Level 2) | C4 Container  | `C4Container`        |
| Internal components (C4 Level 3)               | C4 Component  | `C4Component`        |
| Request flows with numbered steps              | C4 Dynamic    | `C4Dynamic`          |
| Infrastructure, cloud deployment               | C4 Deployment | `C4Deployment`       |
| Cloud services with icons (AWS/GCP/Azure)      | Architecture  | `architecture-beta`  |
| Project timeline, scheduling                   | Gantt         | `gantt`              |
| Proportional data, percentages                 | Pie           | `pie`                |
| Brainstorming, hierarchical ideas              | Mindmap       | `mindmap`            |
| Historical events, chronology                  | Timeline      | `timeline`           |
| Branching strategy, git history                | Git Graph     | `gitGraph`           |
| Flow quantities, resource distribution         | Sankey        | `sankey-beta`        |
| X/Y data visualization                         | XY Chart      | `xychart-beta`       |
| Priority matrix, strategic positioning         | Quadrant      | `quadrantChart`      |
| Layout control, grid positioning               | Block         | `block-beta`         |
| Network packets, protocol headers              | Packet        | `packet-beta`        |
| Task boards, kanban workflow                   | Kanban        | `kanban`             |
| User experience, satisfaction scoring          | User Journey  | `journey`            |
| System requirements traceability               | Requirement   | `requirementDiagram` |

If the user's description doesn't clearly map to one type, suggest 2-3 options with a brief rationale for each, then let them choose.

### When to Load References

Load reference files ONLY when needed for the specific diagram type:

- **C4 diagrams** → Read `references/c4-architecture.md` BEFORE writing code
- **AWS/Cloud architecture** → Read `references/aws-architecture.md` BEFORE writing code
- **Code-to-diagram** → Read `references/code-to-diagram.md` BEFORE analyzing
- **Theming/styling** → Read `references/themes.md` when user requests custom themes
- **Syntax errors** → Read `references/troubleshooting.md` when validation fails
- **Any diagram type details** → Read `references/diagram-types.md` for comprehensive syntax

## Step 2: Create the Diagram

### 2.1 — Write Mermaid Code

Follow these principles in order of priority:

1. **Elegance first** — every diagram must look professional with init directives and curated colors
2. **Correctness** — valid syntax that renders without errors
3. **Clarity** — meaningful labels, logical flow direction, clear relationships
4. **Simplicity** — under 15 nodes per diagram; split complex systems into multiple
5. **Consistency** — uniform naming (camelCase for IDs, descriptive labels in brackets)

### 2.2 — Structure Rules

```
%% Diagram: [Purpose] | Author: [auto] | Date: [auto]
%%{init: {'theme': 'base', 'themeVariables': { ... }}}%%
[diagramType]
    [content]
```

**CRITICAL:** The `%%{init}` directive MUST go on the very first non-comment line, BEFORE the diagram type declaration. Alternatively, use YAML frontmatter at the absolute start of the file.

**Naming conventions:**

- Node IDs: camelCase, descriptive (`orderService`, not `s1`)
- Labels: natural language in brackets (`[Order Service]`)
- Relationships: action verbs (`"Sends order to"`, `"Reads from"`)

**Layout best practices:**

- `TD` (top-down) for hierarchical flows and processes
- `LR` (left-right) for timelines, pipelines, and sequential processes
- `RL` for right-to-left reading contexts
- Use `subgraph` to group related nodes; name subgraphs meaningfully
- Add `direction` inside subgraphs when needed for different flow

### 2.3 — Quick Reference Examples

**Flowchart (with professional styling):**

```mermaid
%%{init: {'theme': 'base', 'themeVariables': {
  'primaryColor': '#4f46e5', 'primaryTextColor': '#fff',
  'primaryBorderColor': '#3730a3', 'lineColor': '#94a3b8',
  'secondaryColor': '#10b981', 'tertiaryColor': '#f59e0b',
  'background': '#ffffff', 'mainBkg': '#f8fafc',
  'nodeBorder': '#cbd5e1', 'clusterBkg': '#f1f5f9',
  'clusterBorder': '#e2e8f0', 'titleColor': '#1e293b',
  'edgeLabelBackground': '#ffffff'
}}}%%
flowchart TD
    Start([Start]) --> Input[/User Input/]
    Input --> Validate{Valid?}
    Validate -->|Yes| Process[Process Data]
    Validate -->|No| Error[Show Error]
    Error --> Input
    Process --> Save[(Save to DB)]
    Save --> End([End])
```

For sequence diagram and ERD styling examples, read `references/themes.md`.

**C4 Context (with MANDATORY elegant styling):**

```mermaid
C4Context
    title System Context — E-Commerce Platform

    Person(customer, "Customer", "Places orders online")
    System(platform, "E-Commerce Platform", "Handles orders and payments")
    System_Ext(payment, "Payment Gateway", "Processes transactions")
    System_Ext(email, "Email Service", "Sends notifications")

    Rel(customer, platform, "Uses", "HTTPS")
    Rel(platform, payment, "Processes payments", "API")
    Rel(platform, email, "Sends emails", "SMTP")

    UpdateRelStyle(customer, platform, $textColor="#475569", $lineColor="#94a3b8")
    UpdateRelStyle(platform, payment, $textColor="#475569", $lineColor="#94a3b8")
    UpdateRelStyle(platform, email, $textColor="#475569", $lineColor="#94a3b8")

    UpdateLayoutConfig($c4ShapeInRow="3", $c4BoundaryInRow="1")
```

**Architecture (AWS with Iconify icons):**

```mermaid
architecture-beta
    group vpc(logos:aws-vpc)[VPC]

    service api(logos:aws-api-gateway)[API Gateway] in vpc
    service lambda(logos:aws-lambda)[Lambda] in vpc
    service db(logos:aws-dynamodb)[DynamoDB] in vpc
    service s3(logos:aws-s3)[S3 Bucket]

    api:R --> L:lambda
    lambda:R --> L:db
    lambda:B --> T:s3
```

**IMPORTANT:** Architecture-beta diagrams with `logos:*` icons require icon pack registration. When rendering with the render script, use the `--icons logos` flag. If rendering in a markdown viewer that doesn't support icon packs, use the built-in icons (`cloud`, `database`, `disk`, `server`, `internet`) as fallback. Read `references/aws-architecture.md` for the complete icon catalog and rendering instructions.

For comprehensive syntax of ALL diagram types, read `references/diagram-types.md`.

## C4 Diagrams — Mandatory Styling Guide

C4 diagrams have **fixed element styling** (blue boxes for systems, gray for persons, etc.), but their **relationship lines default to harsh black** which creates visual noise. You MUST apply these styling rules:

### The C4 Styling Pattern

Every C4 diagram MUST include these directives at the end:

```
    %% === MANDATORY STYLING ===
    %% Apply soft line colors to ALL relationships
    UpdateRelStyle(fromAlias, toAlias, $textColor="#475569", $lineColor="#94a3b8")
    %% Repeat for each Rel() in the diagram

    %% Optimize layout spacing
    UpdateLayoutConfig($c4ShapeInRow="3", $c4BoundaryInRow="1")
```

### C4 Color Values Reference

| Purpose           | Color     | Hex       | Notes                                    |
| ----------------- | --------- | --------- | ---------------------------------------- |
| Soft line color   | Slate-400 | `#94a3b8` | Replaces harsh default black             |
| Line text color   | Slate-600 | `#475569` | Readable but not dominant                |
| Accent line       | Blue-400  | `#60a5fa` | For highlighted or primary relationships |
| Warning line      | Amber-500 | `#f59e0b` | For external/risky connections           |
| Custom element bg | Emerald   | `#10b981` | For data stores or success highlights    |
| Custom element bg | Indigo    | `#4f46e5` | For primary system emphasis              |

### C4 Layout Tips

**CRITICAL — Maximum 6 `Rel()` per diagram.** More than 6 relationships causes Dagre to route arrows through nodes, creating unreadable spaghetti. If your system needs more connections, split it into multiple focused diagrams.

- Use `$c4ShapeInRow="3"` for most diagrams (prevents horizontal crowding)
- Use `$c4ShapeInRow="2"` for diagrams with long labels
- Use `$c4BoundaryInRow="1"` always (stacks boundaries vertically for clarity)
- Apply `$offsetY="-10"` to `UpdateRelStyle` when labels overlap with elements
- Prefer tree-shaped topologies (1 in, 1-2 out per node) over mesh topologies
- Declare elements in flow order — the order of `Container()` declarations affects layout
- Use directional `Rel_D`, `Rel_R`, etc. only as a last resort when auto-layout creates overlapping

For comprehensive C4 syntax, examples, and patterns, read `references/c4-architecture.md`.

## Step 3: Validate

Before rendering, ALWAYS validate the Mermaid syntax:

```bash
node $SKILL_DIR/scripts/validate.mjs <file.mmd>
```

If validation fails:

1. Read the error message carefully
2. Consult `references/troubleshooting.md` for common fixes
3. Fix the syntax and re-validate
4. Maximum 3 fix attempts before asking the user for clarification

## Step 4: Render

### 4.1 — Setup (First Run Only)

```bash
bash $SKILL_DIR/scripts/setup.sh
```

This auto-installs both rendering engines and icon pack dependencies. Run once per environment.

### 4.2 — Single Diagram Rendering

```bash
# SVG (default — best quality)
node $SKILL_DIR/scripts/render.mjs -i diagram.mmd -o diagram.svg

# PNG (for documents/presentations)
node $SKILL_DIR/scripts/render.mjs -i diagram.mmd -o diagram.png --width 1200

# ASCII (for terminals/READMEs)
node $SKILL_DIR/scripts/render-ascii.mjs -i diagram.mmd

# With icon packs (architecture-beta with AWS/tech icons)
node $SKILL_DIR/scripts/render.mjs -i diagram.mmd -o diagram.svg --icons logos,fa
```

The `--icons` flag registers Iconify packs before rendering. Packs: `logos` (AWS/tech), `fa` (Font Awesome). Use `logos` for AWS.

### 4.3 — Batch Rendering

For multiple diagrams at once:

```bash
node $SKILL_DIR/scripts/batch.mjs \
  --input-dir ./diagrams \
  --output-dir ./rendered \
  --format svg \
  --theme default \
  --workers 4
```

### 4.4 — Available Themes

**beautiful-mermaid (15):** `tokyo-night` | `tokyo-night-storm` | `tokyo-night-light` | `dracula` | `nord` | `nord-light` | `catppuccin-mocha` | `catppuccin-latte` | `github-dark` | `github-light` | `solarized-dark` | `solarized-light` | `one-dark` | `zinc-dark` | `zinc-light`

**mermaid-cli native (5):** `default` | `forest` | `dark` | `neutral` | `base`

Custom theme: `--theme base --config '{"theme":"base","themeVariables":{"primaryColor":"#4f46e5","lineColor":"#94a3b8"}}'`

For the full theme catalog, read `references/themes.md`. The render script auto-selects the best engine (mmdc primary, beautiful-mermaid fallback, Puppeteer for icon packs).

## Step 5: Code-to-Diagram (When Requested)

When the user asks to visualize existing code or architecture:

1. Read `references/code-to-diagram.md` for the analysis methodology
2. Analyze the codebase to identify the right diagram type:
   - Module dependencies → Flowchart or Class diagram
   - API routes and handlers → Sequence diagram
   - Database models/schemas → ERD
   - Service architecture → C4 Container or Architecture diagram
   - State machines in code → State diagram
3. Generate the .mmd file **with proper init directives** (Golden Rule 1)
4. Validate and render as usual

## Troubleshooting Quick Reference

| Symptom               | Likely Cause       | Fix                                                    |
| --------------------- | ------------------ | ------------------------------------------------------ |
| Diagram won't render  | Syntax error       | Run validate.mjs, check brackets/quotes                |
| Labels cut off        | Text too long      | Shorten labels or use line breaks `<br/>`              |
| Layout looks wrong    | Wrong direction    | Try different TD/LR/BT/RL                              |
| Nodes overlap         | Too many nodes     | Split into subgraphs or multiple diagrams              |
| Lines too dark/thick  | No init directive  | Add `%%{init}` with `lineColor: '#94a3b8'`             |
| C4 lines overlapping  | No styling applied | Add `UpdateRelStyle` with offsets to each Rel          |
| AWS icons not showing | No icon pack       | Use `--icons logos` flag or fallback to built-in icons |
| `mmdc` not found      | Not installed      | Run `setup.sh`                                         |
| Theme not applied     | Wrong engine       | beautiful-mermaid themes only work with that engine    |

For comprehensive troubleshooting, read `references/troubleshooting.md`.

## Output Conventions

- Save .mmd source files alongside rendered outputs
- Naming: `{purpose}-{type}.mmd` (e.g., `auth-flow-sequence.mmd`)
- For batch: maintain input filename, change extension
- Always provide both the .mmd source and rendered file to the user