agentleFS
lulufoolulufoo/lulu-drawer/skills/sketch/SKILL.md

sketch

Skill8 starsChanged 10 days agoMITView on GitHub

Draw one Sketch on the public viewer.

28 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use touchdesigner-mcp, ant and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • john-data-chenSkill

    touchdesigner-mcp

    Control a running TouchDesigner instance via twozero MCP — create operators, set parameters, wire connections, execute Python, build real-time visuals. 36 native tools.

    Also in 7 projects that use this

  • bergsideSkill

    ant

    Structured, enterprise-focused design system emphasizing clarity, consistency, and efficiency for data-dense web applications.

    Also in 3 projects that use this

  • bergsideSkill

    artistic

    High-contrast, expressive style with creative typography and bold color choices for visually striking interfaces.

    Also in 3 projects that use this

  • +9 more from bergside

The file

We split this skill into the rules it gives an agent

12 rules. 0 are common, 5 are distinctive, and 7 are very specific to this project.

  • 0

    Common

    Most of the 26 edited versions we read keep these.

  • 5

    Distinctive

    This author's own choices. Worth a look.

  • 7

    Very specific

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

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

    Write one Sketch as an HTML file and publish it on the public viewer.

    line 9
  2. Distinctive:

    Write from the template.

    line 20
  3. Distinctive:

    No network. Put every style, shape, glyph, and script in the fragment.

    line 31
  4. Very specific:

    Avoid filter, backdrop-filter, and large blurred shadows.

    line 39
  5. Distinctive:

    Run no endless loop. Keep script work short: the viewer gives up when the sketch does not report a size in time.

    line 40
  6. Distinctive:

    Keep the element count low.

    line 41
  7. Very specific:

    Use cloud mode only when one of these holds: the user asks to save, keep, or update a cloud diagram, or to search cloud diagrams.

    line 47
  8. Very specific:

    If login fails or times out, use short-link mode and tell the user why.

    line 49
  9. Distinctive:

    If the user pastes a viewer URL, load pointers.

    line 50
  10. Very specific:

    Write the fragment to .cache/sketch/<name>.html.

    line 56
  11. Very specific:

    Run $BOARD_CLI encode-source --file <file> --out .cache/sketch --title <title>.

    line 57
  12. Very specific:

    Run $BOARD_CLI preview-diagram --file <skt>, where <skt> is the file that encode-source returned.

    line 58

“Common” is measured: kept by at least 6 in 10 of the 26 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

sketch

Write one Sketch as an HTML file and publish it on the public viewer. Done when a tool url is shown.

CLI

The shared CLI reference defines $BOARD_CLI.

  1. Load cli before the first $BOARD_CLI call of a session.

Template

Write from the template.

  1. Load template.md and write from it.

Render Environment

The viewer mounts the fragment in its own document. Write for this environment.

  1. The fragment renders in a sandboxed iframe. Board and the fragment cannot reach each other's CSS or scripts. Scripts run when the sketch mounts.
  2. :root, html, and body selectors act on that document (zero margin, transparent). A body background paints the diagram only; an html background is ignored.
  3. The frame takes no pointer input by default, so :hover and drag do nothing. A sketch may use click handlers. The canvas pans and zooms over it.
  4. No network. Put every style, shape, glyph, and script in the fragment.

Performance

The frame is resized by its content and exported as one still image. Write for that.

  1. Draw a still diagram. Add animation only when the user asks for it.
  2. Animate only transform and opacity. A change in size resizes the frame on every step.
  3. Avoid filter, backdrop-filter, and large blurred shadows. They are slow to paint.
  4. Run no endless loop. Keep script work short: the viewer gives up when the sketch does not report a size in time.
  5. Keep the element count low.

Mode

Pick short link or cloud. Short link is the default.

  1. Use cloud mode only when one of these holds: the user asks to save, keep, or update a cloud diagram, or to search cloud diagrams; the user pastes a #b: link; this session already created the diagram in cloud mode. Load cloud flow.
  2. Otherwise use short-link mode.
  3. If login fails or times out, use short-link mode and tell the user why.
  4. If the user pastes a viewer URL, load pointers.

Short link flow

Publish unsigned.

  1. Write the fragment to .cache/sketch/<name>.html. <name> is a short English name for the diagram.
  2. Run $BOARD_CLI encode-source --file <file> --out .cache/sketch --title <title>. Always write a short title that names the diagram.
  3. Run $BOARD_CLI preview-diagram --file <skt>, where <skt> is the file that encode-source returned.
  4. Show only the url field and stop.
  5. To change, edit the HTML file and publish again. That creates a new short url.
  6. To keep, use cloud mode (#b:).

Boundaries

  1. CLI contract: $BOARD_CLI --help.

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.