Skip to main content

character-design

Extract character entities from narrative text. Use when analyzing characters, relationships, psychology, development arcs, voice/mocap data, and character variants.

Quellinformationen

Repository
bivex/loreSystem
Letzte Quellaktivität
21. März 2026 um 15:19
Erkannte Sprache von SKILL.md
Englisch
Sterne
0
Forks
0

Installationsoptionen

Standardmäßig ist der Prompt ausgewählt, der zuerst die Quelle prüft. Sie können zu einem direkten Befehl wechseln oder eine lokale Kopie herunterladen.

Quelldateien prüfen

Lesen Sie SKILL.md und alle von SkillsMP angezeigten Begleitdateien, bevor Sie sich für eine Installation entscheiden.

SKILL.md wird angezeigt

SKILL.md
Quellanweisungen · Schreibgeschützte Vorschau
name
character-design
description
Extract character entities from narrative text. Use when analyzing characters, relationships, psychology, development arcs, voice/mocap data, and character variants.
# character-design Domain skill for character extraction. ## Entity Types | Type | Description | |------|-------------| | `character` | Main character entity with name, role, personality, motivation | | `character_evolution` | Character development arc or growth moment | | `character_profile_entry` | Backstory detail or profile information | | `character_relationship` | Relationship between two characters | | `character_variant` | Alternate version, iteration, or form | | `voice_actor` | Voice acting information | | `motion_capture` | Motion capture performance data | ## Extraction Rules 1. **Identify characters**: Named characters with dialogue/actions, referred-to characters, groups 2. **Extract details**: Name, role, status, personality traits, motivations, goals 3. **Track development**: Growth moments, realizations, motivation changes 4. **Map relationships**: Type (friend, rival, family, romantic), strength, dynamics 5. **Note variants**: Alternate forms, timelines, disguises ## Domain Constraints - `backstory`: minimum 100 characters - `ability.power_level`: integer 1–10 - `combat_stats`: attack, defense, health, speed ≥ 0 - `status`: "active" or "inactive" ## Output Format Write to `entities/narrative.json` (narrative-team file): ```json { "characters": [ { "id": 1, "world_id": 1, "name": "Kira", "backstory": "... minimum 100 characters ...", "status": "active", "abilities": [ { "name": "Flame Mastery", "description": "...", "power_level": 8 } ], "created_at": "2026-02-14T10:00:00+00:00", "updated_at": "2026-02-14T10:00:00+00:00", "version": 1 } ], "character_relationships": [ { "id": 2, "character_from_id": 1, "character_to_id": 3, "relationship_type": "friend", "description": "Strong friendship forged through shared battles", "relationship_level": 60, "is_mutual": true, "created_at": "2026-02-14T10:00:00+00:00", "updated_at": "2026-02-14T10:00:00+00:00", "version": 1 } ], "next_id": 3 } ``` ## Key Considerations - **Uniqueness**: Each character has a unique ID; name variations reference the same ID - **Implicit relationships**: Track both explicit and implied connections - **Cross-references**: If needed, track relationships separately in drafts, but final JSON must be compatible with `LoreData.from_dict`.
Auf GitHub ansehen