J

json-canvas

Verified

by kepano · Updated Apr 16, 2026

opus
3.9Kin7.1Kout
$0.351m 46s

Create and edit JSON Canvas files (.canvas) with nodes, edges, groups, and connections. Use when working with .canvas files, creating visual canvases, mind maps, flowcharts, or when the user mentions Canvas files in Obsidian.

jsoncanvasobsidiandiagrammingvisualizationfile-editing

See It In Action

1 turns·claude
The project codebase is in your current working directory. Explore the files to understand the code before making changes. postmortem is in an hour — update the incident review canvas from my outage notes so leadership has the timeline map today

Let me explore the project to find the relevant files.

I have the full picture. The canvas has 3 timeline nodes (header, detection, initial response) plus the Datadog link. The outage notes have 5 more timeline events, root cause analysis, mitigation steps, and follow-up actions that need to be added.

Canvas updated. Here's what was added from your outage notes:

Timeline (left column) — 5 new nodes completing the flow:

  • Investigation (14:41) — connection pool exhaustion on payment-db-primary
  • Root Cause Identified (14:48) — unthrottled backfill_invoice_idx migration
  • Mitigation (14:52) — killed migration, connections recovering
  • Recovery (14:58) — latency normal, error rate <0.1%
  • All-Clear (15:05) — Slack + PagerDuty resolved (green)

Detail cards (right column) — 3 new nodes linked laterally:

  • Root Cause Analysis (red) — batch_size 50000 vs intended 500, row locks exhausting 50-connection pool, CI gap on migration configs
  • Mitigation Steps (yellow) — pg_terminate_backend, pod restart, Datadog verification, scheduler pause
  • Follow-up Actions (cyan) — 5 action items with checkboxes (CI validation, pool alerting, kill-switch, UTC/PST cron retro, runbook update)

All timeline nodes are connected top-to-bottom with arrows, and the detail cards branch off to the right from their corresponding timeline events. Ready for the postmortem.

Claude Sonnet 4.6

Permissions

ScopeDescription
filesystem:read
filesystem:write

SKILL.md

JSON Canvas Skill

File Structure

A canvas file (.canvas) contains two top-level arrays following the JSON Canvas Spec 1.0:

{
  "nodes": [],
  "edges": []
}
  • nodes (optional): Array of node objects
  • edges (optional): Array of edge objects connecting nodes

Common Workflows

1. Create a New Canvas

  1. Create a .canvas file with the base structure {"nodes": [], "edges": []}
  2. Generate unique 16-character hex IDs for each node (e.g., "6f0ad84f44ce9c17")
  3. Add nodes with required fields: id, type, x, y, width, height
  4. Add edges referencing valid node IDs via fromNode and toNode
  5. Validate: Parse the JSON to confirm it is valid. Verify all fromNode/toNode values exist in the nodes array

2. Add a Node to an Existing Canvas

  1. Read and parse the existing .canvas file
  2. Generate a unique ID that does not collide with existing node or edge IDs
  3. Choose position (x, y) that avoids overlapping existing nodes (leave 50-100px spacing)
  4. Append the new node object to the nodes array
  5. Optionally add edges connecting the new node to existing nodes
  6. Validate: Confirm all IDs are unique and all edge references resolve to existing nodes

3. Connect Two Nodes

  1. Identify the source and target node IDs
  2. Generate a unique edge ID
  3. Set fromNode and toNode to the source and target IDs
  4. Optionally set fromSide/toSide (top, right, bottom, left) for anchor points
  5. Optionally set label for descriptive text on the edge
  6. Append the edge to the edges array
  7. Validate: Confirm both fromNode and toNode reference existing node IDs

4. Edit an Existing Canvas

  1. Read and parse the .canvas file as JSON
  2. Locate the target node or edge by id
  3. Modify the desired attributes (text, position, color, etc.)
  4. Write the updated JSON back to the file
  5. Validate: Re-check all ID uniqueness and edge reference integrity after editing

Nodes

Nodes are objects placed on the canvas. Array order determines z-index: first node = bottom layer, last node = top layer.

Generic Node Attributes

AttributeRequiredTypeDescription
idYesstringUnique 16-char hex identifier
typeYesstringtext, file, link, or group
xYesintegerX position in pixels
yYesintegerY position in pixels
widthYesintegerWidth in pixels
heightYesintegerHeight in pixels
colorNocanvasColorPreset "1"-"6" or hex (e.g., "#FF0000")

Text Nodes

AttributeRequiredTypeDescription
textYesstringPlain text with Markdown syntax
{
  "id": "6f0ad84f44ce9c17",
  "type": "text",
  "x": 0,
  "y": 0,
  "width": 400,
  "height": 200,
  "text": "# Hello World\n\nThis is **Markdown** content."
}

Newline pitfall: Use \n for line breaks in JSON strings. Do not use the literal \\n -- Obsidian renders that as the characters \ and n.

File Nodes

AttributeRequiredTypeDescription
fileYesstringPath to file within the system
subpathNostringLink to heading or block (starts with #)
{
  "id": "a1b2c3d4e5f67890",
  "type": "file",
  "x": 500,
  "y": 0,
  "width": 400,
  "height": 300,
  "file": "Attachments/diagram.png"
}

Link Nodes

AttributeRequiredTypeDescription
urlYesstringExternal URL
{
  "id": "c3d4e5f678901234",
  "type": "link",
  "x": 1000,
  "y": 0,
  "width": 400,
  "height": 200,
  "url": "https://obsidian.md"
}

Group Nodes

Groups are visual containers for organizing other nodes. Position child nodes inside the group's bounds.

AttributeRequiredTypeDescription
labelNostringText label for the group
backgroundNostringPath to background image
backgroundStyleNostringcover, ratio, or repeat
{
  "id": "d4e5f6789012345a",
  "type": "group",
  "x": -50,
  "y": -50,
  "width": 1000,
  "height": 600,
  "label": "Project Overview",
  "color": "4"
}

Edges

Edges connect nodes via fromNode and toNode IDs.

AttributeRequiredTypeDefaultDescription
idYesstring-Unique identifier
fromNodeYesstring-Source node ID
fromSideNostring-top, right, bottom, or left
fromEndNostringnonenone or arrow
toNodeYesstring-Target node ID
toSideNostring-top, right, bottom, or left
toEndNostringarrownone or arrow
colorNocanvasColor-Line color
labelNostring-Text label
{
  "id": "0123456789abcdef",
  "fromNode": "6f0ad84f44ce9c17",
  "fromSide": "right",
  "toNode": "a1b2c3d4e5f67890",
  "toSide": "left",
  "toEnd": "arrow",
  "label": "leads to"
}

Colors

The canvasColor type accepts either a hex string or a preset number:

PresetColor
"1"Red
"2"Orange
"3"Yellow
"4"Green
"5"Cyan
"6"Purple

Preset color values are intentionally undefined -- applications use their own brand colors.

ID Generation

Generate 16-character lowercase hexadecimal strings (64-bit random value):

"6f0ad84f44ce9c17"
"a3b2c1d0e9f8a7b6"

Layout Guidelines

  • Coordinates can be negative (canvas extends infinitely)
  • x increases right, y increases down; position is the top-left corner
  • Space nodes 50-100px apart; leave 20-50px padding inside groups
  • Align to grid (multiples of 10 or 20) for cleaner layouts
Node TypeSuggested WidthSuggested Height
Small text200-30080-150
Medium text300-450150-300
Large text400-600300-500
File preview300-500200-400
Link preview250-400100-200

Validation Checklist

After creating or editing a canvas file, verify:

  1. All id values are unique across both nodes and edges
  2. Every fromNode and toNode references an existing node ID
  3. Required fields are present for each node type (text for text nodes, file for file nodes, url for link nodes)
  4. type is one of: text, file, link, group
  5. fromSide/toSide values are one of: top, right, bottom, left
  6. fromEnd/toEnd values are one of: none, arrow
  7. Color presets are "1" through "6" or valid hex (e.g., "#FF0000")
  8. JSON is valid and parseable

If validation fails, check for duplicate IDs, dangling edge references, or malformed JSON strings (especially unescaped newlines in text content).

Complete Examples

See references/EXAMPLES.md for full canvas examples including mind maps, project boards, research canvases, and flowcharts.

References

FAQ

What does json-canvas do?

Create and edit JSON Canvas files (.canvas) with nodes, edges, groups, and connections. Use when working with .canvas files, creating visual canvases, mind maps, flowcharts, or when the user mentions Canvas files in Obsidian.

When should I use json-canvas?

Use it when you need a repeatable workflow that produces code diff.

What does json-canvas output?

In the evaluated run it produced code diff.

How do I install or invoke json-canvas?

Ask the agent to use this skill when the task matches its documented workflow.

Which agents does json-canvas support?

Agent support is inferred from the source, but not explicitly declared.

What tools, channels, or permissions does json-canvas need?

It uses no extra tools; channels commonly include diff; permissions include filesystem:read, filesystem:write.

Is json-canvas safe to install?

Static analysis marked this skill as medium risk; review side effects and permissions before enabling it.

How is json-canvas different from an MCP or plugin?

A skill packages instructions and workflow conventions; tools, MCP servers, and plugins are dependencies the skill may call during execution.

Does json-canvas outperform not using a skill?

About json-canvas

When to use json-canvas

When you need to create a new .canvas file with nodes and edges. When you want to edit or connect items in an existing JSON Canvas document. When you need to validate canvas structure, IDs, and edge references.

When json-canvas is not the right choice

When you need graphical editing inside a visual canvas UI rather than direct file manipulation. When the task involves unrelated diagram formats instead of JSON Canvas files.

What it produces

Produces code diff.