agentleFS
WhatscheapWhatscheap/igly-canvas-plugin/skills/flowchart/SKILL.md

flowchart

Skill0 starsChanged 13 days agoMITView on GitHub

Draw a flowchart, process map, decision tree, user flow or system diagram on an editable Igly canvas. Use when the user asks to diagram steps, a process, a pipeline, an architecture or "how X works".

7 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

5 rules. 0 are common, 1 is distinctive, and 4 are very specific to this project.

  • 0

    Common

    Most of the 5 edited versions we read keep these.

  • 1

    Distinctive

    This author's own choices. Worth a look.

  • 4

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Start/end: shape: "ellipse".

    line 20
  2. Distinctive:

    Never draw arrows by coordinates between boxes.

    line 30
  3. Very specific:

    Default hand-drawn font suits sketches.

    line 33
  4. Very specific:

    Use at most 3 fills: steps #DBEAFE, decisions #FEF3C7, start/end #DCFCE7.

    line 34
  5. Very specific:

    If they want an image, call export_png.

    line 40

“Common” is measured: kept by at least 6 in 10 of the 5 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Flowchart

Draw the whole chart in one add_elements call, then check and fix it.

1. Plan before drawing

  • List the steps (6 to 14 is ideal; group extra detail into notes).
  • Pick the direction: left to right for a pipeline of 5 or fewer steps per row, top to bottom for longer flows and decision trees.
  • Mark decisions (yes/no questions), start and end.

2. Create the board

create_canvas with a short title (e.g. "Customer onboarding"). Keep the returned canvas_id.

3. Lay out on a grid

  • Box: 240 x 110. Column step 320 (left to right) or row step 190 (top to bottom).
  • Start/end: shape: "ellipse". Steps: "rectangle". Decisions: "diamond", 220 x 140. Inputs/outputs: "parallelogram".
  • Branches of a decision go to the next column/row, offset by one lane (240 + 80).
  • Title: type: "text", font_size: 48, bold, 100 above the first box.
  • Give every element a readable id (start, check_email, approve...).

4. Connect

  • type: "arrow" with from_id / to_id. Arrows attach to the facing edges and follow boxes when they move.
  • Label decision branches with the arrow label ("Yes" / "No").
  • Loops and back-edges (retry, "go back to step 1"): connect with from_id/to_id and add curve ("down", "up", "left" or "right") toward the side with empty space, usually away from the title and other rows ("down" for a loop back under a row). It curves around the boxes instead of crossing them.
  • Optional or future steps: dashed: true on the box and on its arrow.
  • Never draw arrows by coordinates between boxes; always use ids.

5. Style

  • Default hand-drawn font suits sketches. For a clean, presentation look set font_family: "Inter" on every element.
  • Use at most 3 fills: steps #DBEAFE, decisions #FEF3C7, start/end #DCFCE7. Keep text dark (#1E1E1E).
  • Optional sticky notes (type: "sticky") beside a step for caveats.

6. Check and finish

  • get_canvas_summary and look for overlaps; fix with update_elements (dx/dy) or arrange (align_middle, distribute_horizontal).
  • Tell the user they can edit the board directly and open it in the full editor with the canvas open_url.
  • If they want an image, call export_png.

Example (one call)

{"canvas_id":"<id>","elements":[
 {"type":"text","id":"title","x":0,"y":-110,"text":"Customer onboarding","font_size":48,"bold":true},
 {"type":"shape","id":"start","shape":"ellipse","x":0,"y":0,"width":240,"height":110,"text":"Sign up","fill":"#DCFCE7"},
 {"type":"shape","id":"verify","x":320,"y":0,"width":240,"height":110,"text":"Verify email","fill":"#DBEAFE"},
 {"type":"shape","id":"ok","shape":"diamond","x":640,"y":-15,"width":220,"height":140,"text":"Verified?","fill":"#FEF3C7"},
 {"type":"shape","id":"done","shape":"ellipse","x":940,"y":0,"width":240,"height":110,"text":"Welcome tour","fill":"#DCFCE7"},
 {"type":"shape","id":"resend","x":640,"y":220,"width":240,"height":110,"text":"Resend link","fill":"#DBEAFE"},
 {"type":"arrow","id":"a1","from_id":"start","to_id":"verify"},
 {"type":"arrow","id":"a2","from_id":"verify","to_id":"ok"},
 {"type":"arrow","id":"a3","from_id":"ok","to_id":"done","label":"Yes"},
 {"type":"arrow","id":"a4","from_id":"ok","to_id":"resend","label":"No"},
 {"type":"arrow","id":"a5","from_id":"resend","to_id":"verify","curve":"down"}
]}

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.