Skip to main content

goal

Quest objectives that drive narrative — the WANTS of adventure

Informations de source

Dépôt
SimHacker/moollm
Dernière activité de la source
9 février 2026 à 21:17
Langue détectée de SKILL.md
anglais
Étoiles
60
Forks
5

Options d'installation

Le prompt qui vérifie d'abord la source est sélectionné par défaut. Vous pouvez passer à une commande directe ou télécharger une copie locale.

Vérifiez les fichiers source

Lisez SKILL.md et les fichiers associés affichés par SkillsMP avant de décider de l'installer.

Explorateur de fichiers
6 fichiers

Affichage de SKILL.md

SKILL.md
Instructions source · Aperçu en lecture seule
name
goal
description
Quest objectives that drive narrative — the WANTS of adventure
allowed-tools
["read_file","write_file"]
tier
1
protocol
SCHEMA-MECHANISM
tags
["moollm","quest","objective","narrative","game"]
related
["adventure","character","needs","hero-story"]
adversary
aimlessness
# Goal > *"A goal is not always meant to be reached. Often it serves simply as something to aim at."* > — Bruce Lee --- ## What Is It? A **Goal** is a quest objective that drives the adventure forward. It defines what the player is trying to achieve and when they've succeeded (or failed). Goals create **narrative tension** — the gap between current state and desired state. --- ## Schema Mechanism Connection From Gary Drescher's theory: ``` Context → Action → Result ``` Goals define the desired **Result** that motivates **Action**. The adventure is a laboratory for learning causal chains that achieve goals. --- ## Goal Properties | Property | Purpose | |----------|---------| | `id` | Unique identifier | | `name` | Short display name | | `description` | What you're trying to achieve | | `status` | pending, active, complete, failed | | `complete_when` | Natural language condition | | `fail_when` | When goal becomes impossible | | `reward` | What you get on completion | | `progress` | Partial completion tracking | --- ## Completion Conditions Natural language conditions that compile to JS/PY: ```yaml complete_when: "player has the treasure" complete_when_js: "(ctx) => ctx.player.inventory.includes('treasure')" ``` --- ## Rewards Goals can grant various rewards: ```yaml reward: buff: "Hero's Glory" # Temporary effect item: "champion-medal" # Object unlock: "secret-room" # New access narrative: "The kingdom celebrates!" # Story beat ``` --- ## Progress Tracking For multi-step goals: ```yaml progress: collected: 3 needed: 5 ``` --- ## Goal Hierarchies Goals can nest: ```yaml parent: defeat-dragon children: - find-magic-sword - learn-dragon-weakness - reach-dragon-lair ``` Complete all children to complete the parent. --- ## Failure Conditions Goals can fail: ```yaml fail_when: "time reaches midnight AND prisoner is not freed" fail_message: "You were too late." ``` --- ## Examples ### Simple Quest ```yaml goal: id: find-treasure name: "Find the Treasure" complete_when: "player has the treasure" ``` ### Timed Challenge ```yaml goal: id: rescue-mission name: "Save the Prisoner" complete_when: "prisoner is freed" fail_when: "10 turns pass without rescue" priority: urgent ``` ### Collection Quest ```yaml goal: id: gather-keys name: "Collect All Keys" complete_when: "player has 5 enchanted keys" progress: collected: 2 needed: 5 ``` --- ## Related Skills - [adventure](../adventure/) — Contains active goals - [hero-story](../hero-story/) — Narrative structure - [needs](../needs/) — Internal motivations --- ## Protocol Symbol ``` SCHEMA-MECHANISM — Goals drive causal learning ```
Voir sur GitHub