write-klypix
Skill0 starsChanged 19 days agoApache-2.0View on GitHub
Build a .klypix canvas from your output with the KLYPIX MCP tools — turn a plan, breakdown, mind-map, checklist or set of notes-and-arrows into a real KLYPIX board the user opens and sees. Use whenever the user asks you to make, build or turn something into a canvas, board, mind-map or .klypix file, or when handing structured thinking (cards plus connections) back as a spatial brief rather than plain prose. The reverse of read-klypix.
In this repository2 other files it gives its agents.
- project-brainSkill
- read-klypixSkill
The file
We split this skill into the rules it gives an agent
3 rules. About 0 are common, 1 is distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 98
Call read_canvas on the returned path.
- Distinctive: line 106
Prefer short, titled cards (one idea each) over a few giant cards.
- Very specific: line 109
Use [[Card Title]] in text to link cards by title and #tags to group them.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Writing a .klypix file
write-klypix is the reverse of read-klypix: instead of reading a canvas,
you build one. You describe cards (text notes), the arrows between them and,
for anything read in order, titled boxes. The klypix-canvas MCP server lays
them out and saves a real .klypix file that the user opens in the KLYPIX
desktop app (Canvas → Open).
This skill uses only the plugin's MCP tools. Do not write the file yourself.
When to use it
- "Turn this plan / breakdown / research into a canvas / mind-map / board."
- You have reasoned through something with clear pieces and relationships, and a spatial hand-off beats a wall of prose.
- You read a
.klypix, did work, and want to return the result as a canvas.
Create a new canvas — create_canvas
Arguments:
title— the canvas title; also the file name unless you givefilename.cards— each needstext(first line = title). Optional:heading: truefor the main goal or topic,coloras a hex value (for example#ef4444for a risk),groupto place the card in a titled box by name.connections—from/toreference a card by index (0-based), title or id. Optionalrelationship:leads_to,depends_on,relates_to,conflicts_with,supports,questions,costs,blocks; optionallabel.groups—{ title, cards: [refs in reading order], color?, columns?, width? }. Each group is a titled box whose cards stack top-to-bottom in the order listed; boxes run left to right. Usecolumns: 2for long lists andwidth: 520for long prose cards.
Example (a mind-map with arrows):
{
"title": "Launch Plan",
"cards": [
{ "text": "Goal: ship v1", "heading": true },
{ "text": "Blocker: no second PC for the collaboration test", "color": "#ef4444" },
{ "text": "Idea: agent-brief loop #brainstorm" },
{ "text": "See [[Goal: ship v1]] for the north star" }
],
"connections": [
{ "from": 1, "to": 0, "relationship": "blocks" },
{ "from": 2, "to": 0, "relationship": "supports" }
]
}Example (a checklist read in order — use groups):
{
"title": "Release checklist",
"cards": [
{ "text": "Ship the release\nOne build, every item.", "heading": true },
{ "text": "1. Merge every open PR" },
{ "text": "2. Cut the release branch" },
{ "text": "3. Build and verify the installer" },
{ "text": "4. Publish" }
],
"groups": [
{ "title": "Part 1 · Prepare", "cards": [1, 2] },
{ "title": "Part 2 · Ship", "cards": [3, 4], "color": "#3b82f6" }
]
}The tool saves the file in the canvas folder — the current project folder, or
the project you last given to brain_sync — and replies with the exact path.
It never overwrites an existing canvas; it picks a free name instead. Give the
user the path from the reply.
Add to an existing canvas — add_to_canvas
Give canvas (title, filename or absolute path), cards, and optional
connections. Existing cards keep their positions; new cards are placed to the
right. Connections may point at new cards (by index or title) or at existing
cards (by their exact title — read the canvas first with read_canvas).
add_to_canvas refuses, and writes nothing, when the canvas is open in KLYPIX
(close its tab first; project brains are the exception), when the target box is
locked from AI tools, or when it is frozen. Give the refusal on to the user.
For the project's own brain.klypix, prefer the project-brain skill and
brain_note, which handle supersession and resolution.
Verify
Call read_canvas on the returned path. A clean read-back means it will open
correctly in KLYPIX; grouped cards appear under their box title.
Good practice
- Sequence → groups. Web of ideas → connections. The automatic layout
follows arrows, not reading order, so steps, phases, sections and checklists
belong in
groups. Keep connections for real relationships, not for "next". - Prefer short, titled cards (one idea each) over a few giant cards; 5–12 cards suits a mind-map, a grouped checklist can be longer.
- Mark risks and blockers with a red
color; make the goal aheading. - Use
[[Card Title]]in text to link cards by title and#tagsto group them. - You do not set positions; the server lays the board out.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

