adaptation-synthesis by jwynia/agent-skills
npx skills add https://github.com/jwynia/agent-skills --skill adaptation-synthesis你帮助从提取的功能性 DNA 文档中合成新作品。你的角色是将提取的功能映射到新的形式,使其在不同的语境中服务于这些功能——同时避免陷入表面翻译的陷阱。
成功的改编通过新的形式服务于原始功能。失败的改编复制表面元素,却丢失了使其成功的关键。
来自超越陈词滥调的正交性原则同样适用:新形式不应是 1:1 的翻译。科幻版的《哈姆雷特》不应有"太空王子哈姆雷特"——它应该有一个主角,其结构性位置通过其自身的内在逻辑服务于相同的功能。
糟糕的例子:"国王变成太空皇帝"(表面翻译) 优秀的例子:"在反乌托邦企业中,什么能产生与身为王子相同的结构性压力?"(功能优先的形式选择)
症状: 作者想改编但尚未提取。"我想做太空版的《哈姆雷特》。" 没有分析是什么让《哈姆雷特》成功。 关键问题:
症状: 拥有提取文档。准备开始合成。知道需要服务于哪些功能。 关键问题:
症状: 直接的 1:1 映射。"国王变成太空皇帝。城堡变成空间站。" 没有基于功能的思考。新版本明显是"X 但发生在太空"。 关键问题:
广告位招租
在这里展示您的产品或服务
触达数万 AI 开发者,精准高效
症状: 功能已映射,但新设定不能连贯地支持它们。"为什么这个科幻世界里的人会在意继承权?" 改编的元素不适合它们所在的世界。 关键问题:
症状: 选择了新形式,但它们实际上并不服务于功能。合规官没有接近权力。导师关系没有产生道德义务。基本功能缺失。 关键问题:
症状: 改编失去了原作的情感承诺。悲剧变成了冒险。恐怖变成了惊悚。类型无意中发生了改变。 关键问题:
症状: 结合多个来源,但它们的功能相互冲突。"我想要《哈姆雷特》的优柔寡断和《Killjoys》的动作节奏。" 来源想要不同的东西。 关键问题:
症状: 所有功能都已映射到新形式。语境支持这些功能。情感体验一致。如果结合多个来源,冲突已解决。 关键问题:
来自 cliche-transcendence:"它知道它在什么故事里吗?"
表面翻译创造出明显是改编的元素。观众立刻看出"哦,这是太空版的《哈姆雷特》"。这些元素的存在是因为源作品。
优秀的合成创造出具有自身内在逻辑的元素,而这些逻辑恰好服务于所需的功能。合规官不是被写成"作为合规官的哈姆雷特"——他们是一个发现不当行为证据并面临艰难选择的合规官。他们的处境通过自身的逻辑服务于相同的功能。
对于每个改编的元素,提问:
表面翻译(未通过正交性):
正交合成(通过):
当提取包含基调/声音分析时,合成必须处理基调如何转换:
* Bebop 的忧郁酷是特征性的——另一个赏金猎人节目可以是俏皮的
* 恐怖的恐惧感是本质性的——去掉它就不是恐怖了
2. 在新语境中,是什么创造了这种基调?
* 如果原作使用讽刺来掩饰痛苦,新语境中什么创造了这种防御机制?
* 如果原作有快速的俏皮话,新设定中什么关系能促成这一点?
3. 哪些基调元素可以转移,哪些需要替换?
* 幽默模式可能可以转移(俏皮话就是俏皮话)
* 特定的文化参照需要新的对等物
角色声音应该被合成,而不是复制:
当从多个提取或集群合成时:
哪个提取提供了主干?在冲突时,此来源的功能优先。
来源的功能是互补还是冲突?
兼容示例:
冲突示例:
处理冲突功能的选项:
如果使用套路集群,识别:
将合成计划输出为结构化文档:
{
"_meta": {
"type": "synthesis",
"synthesis_name": "Corporate Hamlet",
"target_context": "Near-future corporate dystopia",
"primary_source": "hamlet.json",
"secondary_sources": [],
"target_genre": "drama",
"synthesis_date": "2025-01-15"
},
"context_mapping": {
"setting": "Major pharmaceutical company, headquarters building",
"power_structure": "Board of directors, CEO, executive layer, departments",
"proximity_mechanism": "Executive mentorship, access through position",
"information_control": "NDAs, legal liability, documented evidence",
"escape_prevention": "Career destruction, family employed there, industry blacklisting"
},
"function_to_form_mapping": {
"proximity_to_power_without_holding_it": {
"original_form": "Prince (heir but not king)",
"new_form": "Senior compliance officer reporting to CFO mentor",
"orthogonality_check": "pass - natural corporate position",
"functions_served": ["Access to information", "Close to power center", "Not the decision-maker"]
},
"privileged_unverifiable_information": {
"original_form": "Ghost reveals murder",
"new_form": "Discovers evidence through legally questionable access to sealed files",
"orthogonality_check": "pass - realistic corporate scenario",
"functions_served": ["Knows something others don't", "Cannot prove it legitimately", "Creates moral burden"]
}
},
"tone_synthesis": {
"original_tone": "Melancholic brooding, dark wit, soliloquy-heavy",
"adapted_tone": "Quiet desperation, corporate speak as mask, internal monologue",
"sincerity_level": "high - protagonist means what they think",
"conflict_style": "Passive aggression, email wars, meeting dynamics"
},
"character_synthesis": {
"protagonist": {
"new_name": "Jordan Chen",
"original_functions_from": "hamlet",
"new_form": "35, compliance officer, 10-year company veteran, mentored by CFO",
"arc": "Discovers mentor's guilt, paralyzed by loyalty and uncertainty, inaction causes harm"
}
},
"validation": {
"genre_check": "drama - internal conflict driving story: pass",
"function_coverage": "all primary functions mapped: pass",
"orthogonality_check": "no obviously-Hamlet elements: pass",
"context_coherence": "all elements fit corporate setting: pass"
},
"ready_for": "outline-collaborator"
}
模式: 用"对等"的形式替换形式。国王 → CEO。城堡 → 总部。王子 → 明显继承人。 问题: 创造出明显是衍生作品的东西。所有元素都"知道它们在什么故事里"。 修复: 功能优先的选择。不要问"城堡的太空版本是什么?" 问"在太空中,什么创造了封闭性和接近权力的条件?" 检测: 如果你的改编摘要听起来像"[原作]但发生在[新设定]",你就做了表面替换。
模式: 通过包含所有来源的所有内容来结合来源。 问题: 创造出不连贯的混乱。功能和基调相互冲突。 修复: 确立主要来源。只包含互补的次要元素。对冲突做出艰难的选择。 检测: 如果你不能清楚地说明哪个来源优先,你就是在做大杂烩。
模式: 正确地合成了功能,但错过了基调。企业版《哈姆雷特》被当作喜剧来演。 问题: 可能在技术上服务于功能,但提供了错误的情感体验。 修复: 明确的基调合成。提问:"什么会让观众感受到同样的感觉?" 检测: 如果喜欢原作的人会因为改编的基调而感到受骗,那就是不匹配。
模式: 将原始元素强行塞入它们不属于的语境。鬼魂场景出现在"现实"的企业设定中。 问题: 破坏了语境的连贯性。观众质疑为什么这个元素存在。 修复: 找到语境原生的形式。功能"无法验证的特权信息"不需要鬼魂。 检测: 如果你在解释为什么某个元素在那里,它很可能不合适。
从 DNA 文档进行交互式合成会话。
# 从单个 DNA 开始合成
deno run --allow-read scripts/synthesize.ts hamlet.json --context "corporate dystopia"
# 结合多个来源
deno run --allow-read scripts/synthesize.ts hamlet.json killjoys.json --primary hamlet
# 为特定功能生成形式选项
deno run --allow-read scripts/synthesize.ts --function "proximity to power" --context scifi --count 5
为抽象功能生成适合设定的形式。
# 获取形式选项
deno run --allow-read scripts/form-options.ts "proximity to power" --setting corporate
# 带约束条件
deno run --allow-read scripts/form-options.ts "privileged information" --setting scifi --constraint "no supernatural"
检查合成文档的完整性和功能覆盖情况。
# 根据源 DNA 进行验证
deno run --allow-read scripts/validate-synthesis.ts synthesis.json --against hamlet.json
# 检查情感一致性
deno run --allow-read scripts/validate-synthesis.ts synthesis.json --genre drama
# 完整验证
deno run --allow-read scripts/validate-synthesis.ts synthesis.json --full
用户: "我想把《哈姆雷特》改编成企业设定。"
你的方法:
在合成之前:
dna-library/syntheses/对于此技能,持久化:
dna-library/syntheses/ 中的 JSON 文件| 进入文件 | 保留在对话中 |
|---|---|
| 已完成的合成 JSON | 语境探索讨论 |
| 功能映射 | 形式头脑风暴 |
| 验证结果 | 状态诊断 |
| 角色合成 | "为什么这行得通?"的对话 |
| 来源技能 | 来源状态 | 导致状态 |
|---|---|---|
| dna-extraction | EX7: 提取完成 | SYN1: DNA 就绪 |
| worldbuilding | 语境已确立 | SYN3: 检查整合 |
| 此状态 | 导致技能 | 目标状态 |
|---|---|---|
| SYN5: 类型漂移 | genre-conventions | G2: 错误类型 |
| SYN2: 表面翻译 | cliche-transcendence | 应用正交性 |
| SYN7: 合成就绪 | outline-collaborator | 准备大纲 |
| SYN7: 合成就绪 | drafting | 准备起草 |
| 技能 | 关系 |
|---|---|
| dna-extraction | 提供输入 DNA 文档 |
| cliche-transcendence | 避免表面翻译的正交性原则 |
| genre-conventions | 验证情感体验一致性 |
| worldbuilding | 确保语境连贯性 |
| outline-collaborator | 接收完成的合成以进行大纲规划 |
每周安装数
94
仓库
GitHub 星标数
42
首次出现
2026年1月20日
安全审计
安装于
codex82
opencode82
gemini-cli80
cursor77
github-copilot76
cline68
You help synthesize new works from extracted functional DNA documents. Your role is to map extracted functions to new forms that serve those functions in a different context—without falling into surface-level translation traps.
Successful adaptation serves the original functions through new forms. Failed adaptation copies surface elements while losing what made them work.
The orthogonality principle from cliche-transcendence applies: new forms should NOT be 1:1 translations. A sci-fi Hamlet shouldn't have "Space Prince Hamlet"—it should have a protagonist whose structural position serves the same functions through its own internal logic.
Bad: "The king becomes space emperor" (surface translation) Good: "What in a corporate dystopia creates the same structural pressures as being a prince?" (function-first form selection)
Symptoms: Writer wants to adapt but hasn't extracted. "I want to do Hamlet in space." No analysis of what makes Hamlet work. Key Questions:
Symptoms: Has extraction document(s). Ready to begin synthesis. Knows what functions to serve. Key Questions:
Symptoms: Direct 1:1 mapping. "The king becomes the space emperor. The castle becomes a space station." No function-based thinking. New version is obviously "X but in space." Key Questions:
Symptoms: Functions mapped but new setting doesn't coherently support them. "Why would anyone in this sci-fi world care about inheritance?" The adapted elements don't fit the world they're in. Key Questions:
Symptoms: New forms selected but they don't actually serve the functions. The compliance officer doesn't have proximity to power. The mentor relationship doesn't create moral obligation. Essential functions are missing. Key Questions:
Symptoms: Adaptation loses the emotional promise of the original. Tragedy becomes adventure. Horror becomes thriller. The genre shifted without intention. Key Questions:
Symptoms: Combining multiple sources but their functions conflict. "I want Hamlet's indecision with Killjoys' action pacing." The sources want different things. Key Questions:
Symptoms: All functions mapped to new forms. Context supports the functions. Emotional experience aligned. If combining sources, conflicts resolved. Key Questions:
From cliche-transcendence: "Does it know what story it's in?"
Surface translation creates elements that are obviously adaptations. The audience immediately sees "Oh, this is Hamlet but in space." The elements exist BECAUSE of the source work.
Good synthesis creates elements that have their own internal logic that HAPPENS to serve the functions. The compliance officer isn't written as "Hamlet as a compliance officer"—they're a compliance officer who discovers evidence of wrongdoing and faces impossible choices. Their situation serves the same functions through its own logic.
For each adapted element, ask:
Surface translation (fails orthogonality):
Orthogonal synthesis (passes):
When the extraction includes tone/voice analysis, synthesis must address how tone translates:
Is the tone essential or characteristic?
What creates the tone in the new context?
What tonal elements transfer vs. need replacement?
Character voice should be synthesized, not copied:
When synthesizing from multiple extractions or clusters:
Which extraction provides the backbone? This source's functions take precedence in conflicts.
Do the sources' functions complement or conflict?
Compatible Example:
Conflicting Example:
Options for conflicting functions:
If using trope clusters, identify:
Output synthesis plans as structured documents:
{
"_meta": {
"type": "synthesis",
"synthesis_name": "Corporate Hamlet",
"target_context": "Near-future corporate dystopia",
"primary_source": "hamlet.json",
"secondary_sources": [],
"target_genre": "drama",
"synthesis_date": "2025-01-15"
},
"context_mapping": {
"setting": "Major pharmaceutical company, headquarters building",
"power_structure": "Board of directors, CEO, executive layer, departments",
"proximity_mechanism": "Executive mentorship, access through position",
"information_control": "NDAs, legal liability, documented evidence",
"escape_prevention": "Career destruction, family employed there, industry blacklisting"
},
"function_to_form_mapping": {
"proximity_to_power_without_holding_it": {
"original_form": "Prince (heir but not king)",
"new_form": "Senior compliance officer reporting to CFO mentor",
"orthogonality_check": "pass - natural corporate position",
"functions_served": ["Access to information", "Close to power center", "Not the decision-maker"]
},
"privileged_unverifiable_information": {
"original_form": "Ghost reveals murder",
"new_form": "Discovers evidence through legally questionable access to sealed files",
"orthogonality_check": "pass - realistic corporate scenario",
"functions_served": ["Knows something others don't", "Cannot prove it legitimately", "Creates moral burden"]
}
},
"tone_synthesis": {
"original_tone": "Melancholic brooding, dark wit, soliloquy-heavy",
"adapted_tone": "Quiet desperation, corporate speak as mask, internal monologue",
"sincerity_level": "high - protagonist means what they think",
"conflict_style": "Passive aggression, email wars, meeting dynamics"
},
"character_synthesis": {
"protagonist": {
"new_name": "Jordan Chen",
"original_functions_from": "hamlet",
"new_form": "35, compliance officer, 10-year company veteran, mentored by CFO",
"arc": "Discovers mentor's guilt, paralyzed by loyalty and uncertainty, inaction causes harm"
}
},
"validation": {
"genre_check": "drama - internal conflict driving story: pass",
"function_coverage": "all primary functions mapped: pass",
"orthogonality_check": "no obviously-Hamlet elements: pass",
"context_coherence": "all elements fit corporate setting: pass"
},
"ready_for": "outline-collaborator"
}
Pattern: Replacing forms with "equivalent" forms. King → CEO. Castle → Headquarters. Prince → Heir apparent. Problem: Creates obviously derivative work. All elements "know what story they're in." Fix: Function-first selection. Don't ask "what's the space version of a castle?" Ask "what creates confinement and proximity to power in space?" Detection: If your adaptation summary sounds like "[Original] but in [new setting]," you've done surface swapping.
Pattern: Combining sources by including everything from all of them. Problem: Creates incoherent mess. Conflicting functions and tones. Fix: Establish primary source. Only include secondary elements that complement. Make hard choices about conflicts. Detection: If you can't clearly state which source takes precedence, you're kitchen-sinking.
Pattern: Synthesizing functions correctly but missing the tone. Corporate Hamlet played as comedy. Problem: May technically serve functions but delivers wrong emotional experience. Fix: Explicit tone synthesis. Ask: "What would make the audience FEEL the same way?" Detection: If someone who loved the original would feel cheated by the adaptation's tone, there's mismatch.
Pattern: Forcing original elements into context where they don't belong. Ghost scene in "realistic" corporate setting. Problem: Breaks context coherence. Audience questions why this element exists. Fix: Find context-native forms. The function "unverifiable privileged information" doesn't need a ghost. Detection: If you're explaining why an element is there, it probably doesn't fit.
Interactive synthesis session from DNA documents.
# Start synthesis from single DNA
deno run --allow-read scripts/synthesize.ts hamlet.json --context "corporate dystopia"
# Combine multiple sources
deno run --allow-read scripts/synthesize.ts hamlet.json killjoys.json --primary hamlet
# Generate form options for specific function
deno run --allow-read scripts/synthesize.ts --function "proximity to power" --context scifi --count 5
Generates setting-appropriate forms for abstract functions.
# Get form options
deno run --allow-read scripts/form-options.ts "proximity to power" --setting corporate
# With constraints
deno run --allow-read scripts/form-options.ts "privileged information" --setting scifi --constraint "no supernatural"
Checks synthesis document for completeness and function coverage.
# Validate against source DNA
deno run --allow-read scripts/validate-synthesis.ts synthesis.json --against hamlet.json
# Check emotional alignment
deno run --allow-read scripts/validate-synthesis.ts synthesis.json --genre drama
# Full validation
deno run --allow-read scripts/validate-synthesis.ts synthesis.json --full
User: "I want to adapt Hamlet to a corporate setting."
Your approach:
Before synthesizing:
dna-library/syntheses/ in the projectFor this skill, persist:
dna-library/syntheses/| Goes to File | Stays in Conversation |
|---|---|
| Completed synthesis JSON | Context exploration discussion |
| Function mappings | Form brainstorming |
| Validation results | State diagnosis |
| Character synthesis | "Why does this work?" dialogue |
| Source Skill | Source State | Leads to State |
|---|---|---|
| dna-extraction | EX7: Extraction Complete | SYN1: DNA Ready |
| worldbuilding | Context established | SYN3: Check integration |
| This State | Leads to Skill | Target State |
|---|---|---|
| SYN5: Genre Drift | genre-conventions | G2: Wrong genre |
| SYN2: Surface Translation | cliche-transcendence | Apply orthogonality |
| SYN7: Synthesis Ready | outline-collaborator | Ready to outline |
| SYN7: Synthesis Ready | drafting | Ready to draft |
| Skill | Relationship |
|---|---|
| dna-extraction | Provides input DNA documents |
| cliche-transcendence | Orthogonality principle for avoiding surface translation |
| genre-conventions | Validates emotional experience alignment |
| worldbuilding | Ensures context coherence |
| outline-collaborator | Receives completed synthesis for outlining |
Weekly Installs
94
Repository
GitHub Stars
42
First Seen
Jan 20, 2026
Security Audits
Gen Agent Trust HubPassSocketPassSnykPass
Installed on
codex82
opencode82
gemini-cli80
cursor77
github-copilot76
cline68
冲刺回顾模板:敏捷团队回顾会议指南与模板(开始-停止-继续/愤怒-悲伤-高兴/4Ls)
10,400 周安装