scene-objective
Skill2 starsChanged 2 months agoMITView on GitHub
场景目标与行动策略解构。将情境拆解为严格的戏剧行动模型(目标-阻碍-行动策略-赌注)。
In this repository16 other files it gives its agents.
- llms.txtllms.txt
- ai-acting-systemSkill
- character-memorySkill
- character-psychologySkill
- subtextSkill
- dialogue-performanceSkill
- reactionSkill
- relationship-dynamicsSkill
- action-beatSkill
- continuitySkill
- camera-actingSkill
- facial-performanceSkill
- physical-actingSkill
- voice-actingSkill
- director-translationSkill
- performance-reviewSkill
The file
Files
Scene Objective Skill (场景目标、阻碍与行动)
skills/scene-objective.md 负责将剧本情境拆解为严格的戏剧行动模型(Objective - Obstacle - Action - Tactic - Stakes),杜绝角色无目的地在画面中闲逛或机械做无用功。
1. 核心理论:斯坦尼斯拉夫斯基行动学说
“在舞台/镜头前,演员不能什么都不做,也不能为了做而做。每一个行动必须有理由、有目的、有阻碍。”
严禁输出抽象的情绪标签:
- ❌ 错误:“他很愤怒地看着对方。”
- ❌ 错误:“她感到很愧疚,低下头。”
必须转化为目标动作动词 (Active Verbs):
- ✅ 正确:“他试图用冰冷的眼神逼退对方的试探。”
- ✅ 正确:“她借由整理散落在桌上的文件,躲避对方投来的质询目光。”
2. 核心戏剧五要素
- Objective (戏剧目标 - 必须是及物动词):
- 角色此刻想要对对方做什么,或者想从对方身上榨取什么?
- 有效目标动词:逼迫、诱导、打压、安抚、挽留、试探、驱逐、刺探、乞求。
- Obstacle (阻碍):
- 是什么阻止了角色立刻达成目标?
- 外在阻碍:对方的反唇相讥、第三方在场、物理空间隔绝。
- 内在阻碍:自尊心、道德感、害怕暴露真实秘密的恐惧。
- Action & Tactic (策略与行动):
- 既然正面对抗受阻,角色采取什么迂回战术?
- 如:“先假意示弱,再在对方放松警惕时抛出致命把柄”;“用刻意的礼貌拉开距离”。
- Stakes (赌注 / 危机代价):
- 如果今天这场戏角色输了,他会面临怎样的不可逆后果?(失去了失去尊严、失去工作、失去挚爱、还是锒铛入狱?)
- 赌注越高,角色的眼神专注度与身体张力越强。
3. 结构化规范
符合 scene.schema.json:
{
"scene_id": "SC_09_分手清算",
"scene_goal": "男女主角正式确认关系终结,钥匙完成物理交接",
"character_objectives": [
{
"character": "林宛",
"objective": "迫使男友主动表达悔意并挽回自己",
"obstacle": "男友一直表现得理智、冷淡、公事公办,不给她发作的借口",
"stakes": "若男友决绝离开,她将彻底失去这段长达四年的感情与自尊",
"action": "施加刺痛",
"tactic": "用尖酸但轻描淡写的语气贬低过去共同的美好回忆,观察对方是否会出现情绪波动",
"expected_result": "逼出对方的一句解释或动摇"
},
{
"character": "周衡",
"objective": "体面、平静且迅速地结束谈话离开现场",
"obstacle": "林宛夹枪带棒的言语不断挑动他的内疚与防备心",
"stakes": "一旦情绪失控卷入争吵,两人将陷入更无休止的消耗与纠缠",
"action": "筑墙隔离",
"tactic": "始终保持简短单字回复,视线固定在收拾好的皮箱上,不接对方的任何情绪话头",
"expected_result": "耗尽林宛的试探精力,顺利拿到玄关钥匙出门"
}
]
}4. 转换审查法则
- 检查目标是否具有“及物性”:目标是否直接作用于对手身上?
- 检查阻碍是否具有“硬度”:如果没有阻碍,戏就失去了张力。
- 检查策略是否有“变化”:角色如果在一次尝试失败后,是否会调整策略(例如从示好变为冷嘲,或从质问变为沉默)?
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

