Menú
Add a complete Repo Dungeon biome content pack, including tileset references, prop/NPC dressing, palette cues, and ambient audio hooks for a language or topic zone.
Instalar con Codex o Claude Copia este prompt, pégalo en Codex, Claude u otro asistente, y deja que revise la página de la skill y la instale por ti.
Basado en la clasificación ocupacional SOC
| name | add-biome-content-pack |
| description | Add a complete Repo Dungeon biome content pack, including tileset references, prop/NPC dressing, palette cues, and ambient audio hooks for a language or topic zone. |
Use this skill when introducing or expanding a biome so its assets, props, and audio remain consistent across Repo Dungeon.
Start from the language/topic mapping and determine the biome ID, palette, room dressing expectations, and any special props or NPC flavor.
Confirm which tilesets, sprites, loot art, and audio cues are needed. Note any placeholder or licensed assets that must be documented.
Attach the asset pack to biome configuration, room props, NPC presentation, and any tutorial or map-facing labels.
export const biomeContent = {
id: 'example-biome',
tilesetKey: 'example-tileset',
ambientTrackKey: 'example-ambient',
};
Provide mute-safe audio hooks, reduced-motion fallbacks, and bundle-conscious asset loading behavior.
Test the biome inside a generated zone to confirm visuals, interactions, and audio remain coherent with the rest of the dungeon.
See docs/PRD.md for the full specification:
Add Repo Dungeon unit, component, or integration coverage for a gameplay, UI, or platform subsystem using Vitest, React Testing Library, and MSW as appropriate.
Scaffold a Repo Dungeon Phaser scene with lifecycle hooks, typed event wiring, and integration points for generation, UI overlays, and progression events.
Build an accessible Repo Dungeon React overlay, panel, or screen that consumes shared game state without moving feature logic out of its specialist owner.
Add or extend a Repo Dungeon GitHub data flow, including Octokit access, typed models, caching, backoff, and mocked verification coverage.
Discover which LLM models are available to the user (cloud subscription + local Ollama), classify each generated agent's workload, and recommend (and optionally apply) a per-agent model assignment so that lightweight agents do not default to the most expensive model. Use this skill after `forge-build-agent-team` has produced an agent team, or any time the team changes.
Meta-skill that bootstraps a new project end-to-end from a one-liner idea by chaining the deterministic Forge skills: forge-build-prd → (pause for review) → forge-build-agent-team → (pause for review) → optionally forge-assign-models. Use this skill when a user wants to go from idea to a reviewable PRD and a generated agent team in a single guided flow, without losing the human review gates between steps.