anbeime/skill / skills/hairline/hairline/skills/hairline-create
hairline-create
VerifiedUse when someone asks for a new Hairline figure, or runs /hairline-create with an idea. Draws one isometric line figure that answers the pointer, in the style and on the engine of @lucasmarkes/hairline, and hands it over as a single self-contained HTML file.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 7.0 KB |
| bench.html | other | 8.1 KB |
| build.mjs | script | 2.1 KB |
| concepts.md | markdown | 6.3 KB |
| examples/riffle.js | script | 7.2 KB |
| examples/terrain.js | script | 4.3 KB |
| kernel.js | script | 27.3 KB |
| look.md | markdown | 9.7 KB |
| look.mjs | script | 28.0 KB |
| read.mjs | script | 19.2 KB |
| rules.md | markdown | 10.3 KB |
| source.md | markdown | 3.3 KB |
| validate.mjs | script | 15.3 KB |
source.md
View on GitHub# From a drawing the person already has
`read.mjs` wrote two files: `hairline-<name>-source.png`, the SVG as it looks (on dark when its ink is light, and the card's ink line says so), and `hairline-<name>-source.txt`, the card: its viewBox, title and ids, the ink's box, and one line per painted shape with its box, centre, paint and its relation to the shape it sits in or touches. A path drawn in several pieces, a wordmark's letters in one `<path>`, gives one line per piece, each marked `part 2 of 9`. Look at the picture. Read the card. Then answer these, in order, before any concept is written.
**The SVG gives the identity, not the paths. The figure is rebuilt as solids, with the source's proportions, in the kernel's style. Never trace a path.** The person will know their object by its parts and its proportions, not by its coordinates.
1. **What is the object?** One word, from the picture, before reading the card. Then check it against the `title`, the ids and the card. When the picture and the card disagree, the picture wins: ids are what the author typed, the picture is what they drew.
2. **Which parts make it that object?** Two or three, each pointing at card lines: `#1 the body, #3 and #4 the fins, #5 the flame`. This is step 2 of `concepts.md`, "find what gives it away", now with evidence. Keep the list: the figure draws from it, and the look (`look.md`, item 14) checks the rest pose against it.
3. **Which proportions must survive?** The ink box's ratio, and the two or three relations between parts that weigh most: the fins at the base, each as wide as a third of the body; the flame below, a third of the body's height. These are the numbers that go into `rings` and `prism`. Everything else is yours to decide in the kernel's terms.
4. **What is decoration?** Shadows, glows, double outlines, the small shapes the card counted, anything that is there for colour. It stays out. A figure has one stroke in four weights (rule 04): what the SVG said with colour, the figure says with one bright mark.
5. **What does the object answer?** If the request named the gesture, it is fixed: the concepts keep it and differ in the form the object takes and in what the read-out says. If it did not, ask what a hand would do with this object. The table "Six answers already proven" in `concepts.md` is the menu, and each concept takes a different answer from it.
Then write the concepts as step 1 of `SKILL.md` says, all with the same object and the same parts.
## What the card cannot say
- **Depth.** The SVG is flat. You decide what stands up and what lies down: a rocket stands on its fins, a card lies on its tray. The rest pose (rule 05) is where that decision shows.
- **Which part answers.** The card lists parts; it does not know which one a hand would reach for. Prefer the concept whose answering part is also the one that gives the object away.
- **Whose it is.** A mark that belongs to someone else is theirs to publish or not: say so at hand-over, as "From a mark" in `concepts.md` says.
## Without a browser
The card is reduced to the viewBox, the title, the ids, the counts and what was not read, and there is no picture. Read the SVG's own text for the rest: each shape's attributes give its box well enough for proportions. Say at hand-over, in these words, that the source was **not seen in a browser**.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |