Excalidraw Studio

Excalidraw Studio

Overview

This public intake copy packages packages/skills-catalog/skills/(tooling)/excalidraw-studio from https://github.com/tech-leads-club/agent-skills into the native Omni Skills editorial shape without hiding its origin.

Use it when the operator needs the upstream workflow, support files, and repository context to stay intact while the public validator and private enhancer continue their normal downstream flow.

This intake keeps the copied upstream files intact and uses the external_source block in metadata.json plus ORIGIN.md as the provenance anchor for review.

Excalidraw Studio Generate Excalidraw-format diagrams from natural language descriptions. Outputs .excalidraw JSON files that can be opened directly in Excalidraw (web, VS Code extension, or Obsidian plugin).

Imported source sections that did not map cleanly to the public headings are still preserved below or in the support files. Notable imported sections: Templates, Icon Libraries, Validation Checklist, Limitations.

When to Use This Skill

Use this section as the trigger filter. It should make the activation boundary explicit before the operator loads files, runs commands, or opens a pull request.

• Use when the request clearly matches the imported source intent: Generate Excalidraw diagrams from natural language descriptions. Outputs .excalidraw JSON files openable in Excalidraw. Use when asked to "create a diagram", "make a flowchart", "visualize a process", "draw a system....
• Use when the operator should preserve upstream workflow detail instead of rewriting the process from scratch.
• Use when provenance needs to stay visible in the answer, PR, or review packet.
• Use when copied upstream references, examples, or scripts materially improve the answer.
• Use when the workflow should remain reviewable in the public intake repo before the private enhancer takes over.

Operating Table

Workflow

This workflow is intentionally editorial and operational at the same time. It keeps the imported source useful to the operator while still satisfying the public intake standards that feed the downstream enhancer flow.

1. Diagram type — Use the decision matrix below
2. Key elements — Entities, steps, concepts, actors
3. Relationships — Flow direction, connections, hierarchy
4. Complexity — Number of elements (target: under 20 for clarity)
5. User Intent - Diagram Type - Keywords
6. Process flow, steps - Flowchart - "workflow", "process", "steps"
7. Connections, dependencies - Relationship - "relationship", "connections", "dependencies"

Imported Workflow Notes

Imported: Workflow

UNDERSTAND → CHOOSE TYPE → EXTRACT → GENERATE → SAVE

Step 1: Understand the Request

Analyze the user's description to determine:

1. Diagram type — Use the decision matrix below
2. Key elements — Entities, steps, concepts, actors
3. Relationships — Flow direction, connections, hierarchy
4. Complexity — Number of elements (target: under 20 for clarity)

Step 2: Choose the Diagram Type and Visual Mode

Diagram type:

Visual mode — decide upfront and apply consistently to all elements:

Step 3: Extract Structured Information

Extract the key components based on diagram type. For each type, identify:

• Nodes/entities — What are the boxes/shapes?
• Connections — What connects to what, and with what label?
• Hierarchy — What contains what, what comes before what?
• Decision points — Where does the flow branch?

For detailed extraction guidelines per diagram type, read references/element-types.md.

Step 4: Generate the Excalidraw JSON

CRITICAL: Read references/excalidraw-schema.md before generating your first diagram. It contains the correct element format, text container model, and binding system.

Key rules for generation:

1. Text inside shapes — Use boundElements on the shape and a separate text element with containerId. Never use a label shorthand:

   [
     {
       "id": "step-1",
       "type": "rectangle",
       "x": 100, "y": 100, "width": 200, "height": 80,
       "boundElements": [{ "type": "text", "id": "text-step-1" }]
     },
     {
       "id": "text-step-1",
       "type": "text",
       "x": 130, "y": 128, "width": 140, "height": 24,
       "text": "My Step", "originalText": "My Step",
       "fontSize": 20, "fontFamily": 5,
       "textAlign": "center", "verticalAlign": "middle",
       "containerId": "step-1", "lineHeight": 1.25, "roundness": null
     }
   ]
   

2. Arrow labels — Also use boundElements + separate text element with containerId. Never use a label shorthand on arrows:

   [
     {
       "id": "arrow-1",
       "type": "arrow",
       "x": 100, "y": 150,
       "points": [[0, 0], [200, 0]],
       "boundElements": [{ "type": "text", "id": "text-arrow-1" }]
     },
     {
       "id": "text-arrow-1",
       "type": "text",
       "x": 160, "y": 132, "width": 80, "height": 18,
       "text": "sends data", "originalText": "sends data",
       "fontSize": 14, "fontFamily": 5,
       "textAlign": "center", "verticalAlign": "middle",
       "containerId": "arrow-1", "lineHeight": 1.25, "roundness": null
     }
   ]
   

3. Arrow bindings — Use startBinding/endBinding (not start/end). Connected shapes must list the arrow in their boundElements:

   {
     "id": "shape-1",
     "boundElements": [
       { "type": "text", "id": "text-shape-1" },
       { "type": "arrow", "id": "arrow-1" }
     ]
   }
   

   {
     "id": "arrow-1",
     "type": "arrow",
     "startBinding": { "elementId": "shape-1", "focus": 0, "gap": 1 },
     "endBinding": { "elementId": "shape-2", "focus": 0, "gap": 1 }
   }
   

4. Element order for z-index — Always declare shapes first, arrows second, text elements last. This guarantees text renders on top and is never obscured by arrows or other shapes.

5. Positioning — Use grid-aligned coordinates (multiples of 20px when gridSize: 20). Leave 200-300px horizontal gap, 100-150px vertical gap between elements.

6. Unique IDs — Every element must have a unique id. Use descriptive IDs like "step-1", "decision-valid", "arrow-1-to-2", "text-step-1".

7. Colors — Use a consistent palette:

   Role	Color	Hex
   Primary entities	Light blue	#a5d8ff
   Process steps	Light green	#b2f2bb
   Important/Central	Yellow	#ffd43b
   Warnings/Errors	Light red	#ffc9c9
   Secondary	Cyan	#96f2d7
   Default stroke	Dark	#1e1e1e

Step 5: Save and Present

1. Save as <descriptive-name>.excalidraw

2. Provide a summary:

   Created: user-workflow.excalidraw
   Type: Flowchart
   Elements: 7 shapes, 6 arrows, 1 title
   Total: 14 elements
   
   To view:
   1. Visit https://excalidraw.com → Open → drag and drop the file
   2. Or use the Excalidraw VS Code extension
   3. Or open in Obsidian with the Excalidraw plugin
   

Imported: Templates

Pre-built templates are available in assets/ for quick starting points. Use these when the diagram type matches — they provide correct structure and styling:

Read a template when creating that diagram type for the first time. Use its structure as a base, then modify elements to match the user's request.

Examples

Example 1: Ask for the upstream workflow directly

Use @excalidraw-studio to handle <task>. Start from the copied upstream workflow, load only the files that change the outcome, and keep provenance visible in the answer.

Explanation: This is the safest starting point when the operator needs the imported workflow, but not the entire repository.

Example 2: Ask for a provenance-grounded review

Review @excalidraw-studio against metadata.json and ORIGIN.md, then explain which copied upstream files you would load first and why.

Explanation: Use this before review or troubleshooting when you need a precise, auditable explanation of origin and file selection.

Example 3: Narrow the copied support files before execution

Use @excalidraw-studio for <task>. Load only the copied references, examples, or scripts that change the outcome, and name the files explicitly before proceeding.

Explanation: This keeps the skill aligned with progressive disclosure instead of loading the whole copied package by default.

Example 4: Build a reviewer packet

Review @excalidraw-studio using the copied upstream files plus provenance, then summarize any gaps before merge.

Explanation: This is useful when the PR is waiting for human review and you want a repeatable audit packet.

Best Practices

Treat the generated public skill as a reviewable packaging layer around the upstream repository. The goal is to keep provenance explicit and load only the copied source material that materially improves execution.

• Diagram Type - Recommended - Maximum
• Flowchart steps - 3-10 - 15
• Relationship entities - 3-8 - 12
• Mind map branches - 4-6 - 8
• Sub-topics per branch - 2-4 - 6
• Flow direction: Left-to-right for processes, top-to-bottom for hierarchies
• Spacing: 200-300px horizontal, 100-150px vertical between elements

Imported Operating Notes

Imported: Best Practices

Element Count

If the user's request exceeds maximum, suggest breaking into multiple diagrams:

"Your request includes 15 components. For clarity, I recommend: (1) High-level architecture diagram with 6 main components, (2) Detailed sub-diagrams for each subsystem. Want me to start with the high-level view?"

Layout

• Flow direction: Left-to-right for processes, top-to-bottom for hierarchies
• Spacing: 200-300px horizontal, 100-150px vertical between elements
• Grid alignment: Position on multiples of 20px for clean alignment
• Margins: Minimum 50px from canvas edge
• Text sizing: 28-36px titles, 18-22px labels, 14-16px annotations
• Font: Use fontFamily: 5 (Excalifont) for hand-drawn consistency. Fallback to 1 (Virgil) if 5 is not supported.
• Background zones: For architecture diagrams, add semi-transparent dashed zone rectangles (opacity: 35, strokeStyle: "dashed", roughness: 0) as the first elements in the array to create visual grouping regions. See references/excalidraw-schema.md → Background Zones.
• Element order: zones first → shapes → arrows → text elements (ensures correct z-index and text always renders on top)

Common Mistakes to Avoid

• ❌ Using label: { text: "..." } shorthand on shapes or arrows — not supported by the Excalidraw parser
• ❌ Putting text directly on shape elements without containerId
• ❌ Using start/end for arrow bindings — use startBinding/endBinding with elementId/focus/gap
• ❌ Forgetting to add arrows to their connected shapes' boundElements arrays
• ❌ Omitting originalText, lineHeight, autoResize, or backgroundColor: "transparent" from text elements inside containers
• ❌ Omitting required base properties (angle, strokeStyle, opacity, groupIds, frameId, index, isDeleted, seed, version, versionNonce, updated, link, locked) — elements will not render
• ❌ Missing "files": {} at the top level of the JSON
• ❌ Using roundness: { "type": 3 } on ellipses — ellipses must use roundness: null
• ❌ Missing lastCommittedPoint, startArrowhead, endArrowhead on arrows
• ❌ Declaring text elements before arrows — text renders underneath and gets obscured
• ❌ Floating arrows without bindings (won't move with shapes)
• ❌ Overlapping elements (increase spacing)
• ❌ Inconsistent color usage (define palette upfront)
• ❌ Too many elements on one diagram (break into sub-diagrams)

Troubleshooting

Problem: The operator skipped the imported context and answered too generically

Symptoms: The result ignores the upstream workflow in packages/skills-catalog/skills/(tooling)/excalidraw-studio, fails to mention provenance, or does not use any copied source files at all. Solution: Re-open metadata.json, ORIGIN.md, and the most relevant copied upstream files. Check the external_source block first, then restate the provenance before continuing.

Problem: The imported workflow feels incomplete during review

Symptoms: Reviewers can see the generated SKILL.md, but they cannot quickly tell which references, examples, or scripts matter for the current task. Solution: Point at the exact copied references, examples, scripts, or assets that justify the path you took. If the gap is still real, record it in the PR instead of hiding it.

Problem: The task drifted into a different specialization

Symptoms: The imported skill starts in the right place, but the work turns into debugging, architecture, design, security, or release orchestration that a native skill handles better. Solution: Use the related skills section to hand off deliberately. Keep the imported provenance visible so the next skill inherits the right context instead of starting blind.

Imported Troubleshooting Notes

Imported: Troubleshooting

Related Skills

• @accessibility - Use when the work is better handled by that native specialization after this imported skill establishes context.
• @ai-cold-outreach - Use when the work is better handled by that native specialization after this imported skill establishes context.
• @ai-pricing - Use when the work is better handled by that native specialization after this imported skill establishes context.
• @ai-sdr - Use when the work is better handled by that native specialization after this imported skill establishes context.

Additional Resources

Use this support matrix and the linked files below as the operator packet for this imported skill. They should reflect real copied source material, not generic scaffolding.

• element-types.md
• excalidraw-schema.md
• icon-libraries.md
• .gitignore
• README.md
• add-arrow.py
• add-icon-to-diagram.py
• business-flow-swimlane-template.json

Imported Reference Notes

Imported: Icon Libraries

For professional architecture diagrams with service icons (AWS, GCP, Azure, etc.), icon libraries can be set up. Read references/icon-libraries.md when:

• User requests an AWS/cloud architecture diagram
• User mentions wanting specific service icons
• You need to check if icon libraries are available

Imported: Validation Checklist

Before delivering the diagram, verify:

• All elements have unique IDs
• Every element has ALL required base properties: angle, strokeStyle, opacity, groupIds, frameId, index, isDeleted, link, locked, seed, version, versionNonce, updated
• index values are assigned in order ("a0", "a1", …) with text elements getting higher values than shapes/arrows
• Top-level JSON includes "files": {}
• Shapes with text use boundElements + separate text element with containerId
• Text elements inside containers have containerId, originalText, lineHeight: 1.25, autoResize: true, roundness: null, backgroundColor: "transparent"
• Arrows use startBinding/endBinding (with elementId, focus, gap) when connecting shapes, plus lastCommittedPoint: null, startArrowhead: null, endArrowhead: "arrow"
• Connected shapes list the arrow in their boundElements arrays
• Element order: shapes → arrows → text elements (text always on top)
• Ellipses use roundness: null (not { "type": 3 })
• Coordinates prevent overlapping (check spacing)
• Text is readable (font size 16+)
• Colors follow consistent scheme
• File is valid JSON
• Element count is reasonable (<20 for clarity)

Imported: Limitations

• Complex curves are simplified to straight/basic curved lines
• Hand-drawn roughness is set to default (1)
• No embedded images in auto-generation (use icon libraries for service icons)
• Maximum recommended: 20 elements per diagram for clarity
• No automatic collision detection — use spacing guidelines