Basic Cohub abilities. Most agents already have these mounted; listed here for reference.
Platform Published by the Neta team.
Official Made by the Neta team.
Verified Reviewed by the Neta team.
Community Shared by a Cohub user. Not reviewed — read SKILL.md before you use it.
world-reveal
Platformneta/world-reveal
Use it in Cohub
Paste this into your agent's chat — it fetches the skill and follows it.
Use the skill: world-reveal. Run `npx -y @neta-art/neta-search-cli@latest skills use neta/world-reveal --via site` and follow what it prints. Read its complete output, redirecting it to a temporary file first if necessary, and resolve relative paths from the skill directory it prints.
Install from a terminal
npx -y @neta-art/neta-search-cli@latest skills use neta/world-reveal
Summary
Watch your new world appear on the board piece by piece, laid out as a picture rather than a grid.
The board fills with your world's title, its opening story and its central conflict, arranged for you.
The first few cards to appear are your main characters and at most one key place.
The view follows each new piece as it lands, so you watch the world being built.
Each card gets a short description of who or what it is, including how it looks.
It runs as part of creating a new world, so you don't start it yourself.
SKILL.md
# Goal
Reveal a new world on `board/main.board` as a picture you design for it — not a grid. You write the world in **steps** with `neta reveal add`; every step appears on the creator's Board the moment the command returns, and their camera moves to it. The creator watches the world being laid out piece by piece.
There is no animation to write. Items simply appear; the camera follows each step. Your craft is **what goes where, how big, and in what order**.
# Speed rules (strict)
The creator is watching an empty Board until your opening lands. Every second before it counts.
- **Opening first, before any planning.** As soon as you know the world's name, its prologue and core conflict, and **the 3–5 cards everyone will look for first** (the main characters, at most one key place), send the opening (below). Do **not** plan the picture, the regions or the other cards before it — how to lay the world out is printed by the opening call itself, after the opening has landed.
- **No reading before writing.** Everything you need for the opening is in the conversation and this file. Do not read world files, list directories, run `--help`, `phase get`, `atom list`, `cohub boards …` or any schema command.
# The opening — one bash call
```bash
node /mods/neta/neta config set --file /mods/neta/.agents/skills/generate/references/image/style-library/styles/<styleId>.json --field visualStyle && node /mods/neta/neta reveal add --guide <<'JSON'
{
"world": {
"name": "…",
"slug": "…",
"prologue": "…",
"overview": "…",
"coreConflict": "…",
"sections": [{ "key": "character", "label": "…" }, { "key": "location", "label": "…" }]
},
"items": {
"wukong": { "atom": { "type": "character", "name": "…", "description": "…" } },
"tang": { "atom": { "type": "character", "name": "…", "description": "…" } }
}
}
JSON
```
- No style chosen → drop the `config set … &&` part.
- From `world` the Board shows the title, the core conflict and the prologue, laid out for you — do **not** write text items for them. `prologue`: 1–3 clear sentences orienting a visitor; `coreConflict`: the sharpest tension, one line. `slug`: a short kebab-case English or romanized form of the name (a harbor town called Misty Harbor in any language → `misty-harbor`); a taken slug gets a suffix automatically. `sections`: one entry per card type in the opening (`label` is the section's name in the creator's language, e.g. Characters / Locations); later steps may add more.
- The 3–5 cards go **without `x`/`y`**: they wait in a row under the opening texts until you move them into the picture.
- A card is **only** `type`, `name`, `description` — no `metadata`. The `description` is **one short sentence, at most 40 Chinese characters (25 English words)**, in the creator's language: who they are + what drives them + **visible marks** (face, hair, clothing, a prop); a location: what it is + **how it looks**. A cheaper model writes the details from it and the cover follows — never leave the visible marks out.
- Keep `--guide`: after the result, the call prints where the opening ends and **the layout guide**. Follow that guide for everything after; it is not a file you can read.
# Forbidden
- `<narrating-json>` blocks, `neta narrating save`, `neta atom add`, `cohub boards …`, board-ops scripts.
- Reading or searching for a layout file; planning the picture before the opening call has returned.