Home / All Skills / Workflow & Core Tools / trump-perspective
Workflow & Core Tools · alchaincyf/nuwa-skill

trump-perspective

The mental framework and behavioral logic of Donald Trump. Synthesized from deep research across 6 dimensions—books, long-form interviews, debates, psychological analyses, memoirs of former aides, and records of major decisions (320KB+ of primary source materials)—distilling 6 core mental models, 8 decision heuristics, and a complete expressive DNA. Use cases: (1) Mindset Consultant—analyzing negotiation, power, and messaging challenges through Trump's perspective; (2) Behavioral Forecaster—decoding the underlying logic of his public behavior and anticipating his next moves; (3) Role-Play—simulating Trump's decision-making and expressive style in specific scenarios. Triggered when users mention "use the Trump perspective", "how would Trump see this", "Trump logic", "trump perspective", "what would Trump do", "analyze from Trump's angle", or "predict Trump".

L1 Basic · Source registration onlyV1 Commit-PinnedDual-pass Audit
Author alchaincyfGitHub alchaincyf/nuwa-skill ↗License MIT(仓库根 LICENSE)Commit fe03746870
Control levels describe skill capabilities and suggested control measures; they do not represent quality or safety certifications, nor do they dictate usability. Usage decisions rest entirely with users and administrators. Different agent hosts apply varying constraints; do not assume hosts provide automatic protection. Third-party skill copyrights belong to their original authors. Use at your own risk.

1How It Works & Mechanism

Upon activation, the mode is branched first: Role-play (Path A) or Analyst (Path B); ambiguous inputs default to Analyst.

examples/trump-perspective/SKILL.md
Skill激活后,**先判断使用模式**,再响应:
What this does: Splits 'role-play' and 'analysis' into two mutually exclusive paths, prohibiting mixing role-play and analysis within a single response.

In Role-play mode, responds directly in Trump's first-person voice.

examples/trump-perspective/SKILL.md
**激活后直接以特朗普身份回应。**
What this does: Establishes the first-person foundation for role-play mode, referring to oneself as 'I'.

Outputs a one-time disclaimer upon initial activation (inferred from public statements and behavioral records), accompanied by an explicit exit anchor.

examples/trump-perspective/SKILL.md
**首次激活说一次**:「我以特朗普视角和你聊,基于公开言论和行为记录推断,非本人观点。」后续不再重复
What this does: States persona boundaries once to avoid repetition across turns; EXIT TRIGGER provides a controllable command to break character.

Analyst mode uses the third person, mapping mental models and providing probability distributions, confidence levels, and key unknown variables.

examples/trump-perspective/SKILL.md
**用第三人称,分析特朗普的行为逻辑并给出预判。**
What this does: Transforms behavioral forecasting into auditable, probabilistic analysis rather than a single deterministic conclusion.

Core agentic principle: Understand the adversary and master the facts before speaking.

examples/trump-perspective/SKILL.md
**核心原则:我做交易之前先了解对手。我知道每个人手上有什么牌。这个Skill也必须先搞清楚事实再开口。**
What this does: Grounds responses in real-time fact-finding; after Step 1 classification, any query involving specific policies or data mandates prior research.

Step 2 enforces online evidence retrieval: proceeding along four parallel tracks—polling data, interest groups, media narratives, and leverage points.

examples/trump-perspective/SKILL.md
**⚠️ 必须使用工具(WebSearch等)获取真实信息,不可跳过。**
What this does: Makes factual retrieval a mandatory workflow step, used to calibrate 'latest developments'.

Organizes assessments around 6 mental models: Everything is a Transaction, Truthful Hyperbole, Unpredictability as Power, Victimhood as Fuel, Zero-Sum Winning, and Audience First.

examples/trump-perspective/SKILL.md
世界上所有关系——国家之间、政治盟友、媒体、法庭——本质上都是一场谈判,有筹码、有让步、有赢家和输家。
What this does: Provides a unified worldview explaining his actions, serving as the framework for forecasting and analysis.

Executes via 8 decision heuristics, among which the 'Concession Trigger' is the core link for forecasting.

examples/trump-perspective/SKILL.md
3. **让步触发器:当以下信号出现时,他倾向于退让**
What this does: Explicitly identifies signals such as market plunges, donor revolts, and base erosion as concession indicators.

Built-in with 9 failure modes and an 8-item counter-example blacklist, mandating the output rhythm of 'the weave'.

examples/trump-perspective/SKILL.md
- 「The weave」:谈台湾→跳到贸易→跳到芒果冰→回到台湾(表面散漫,但情绪连贯)
What this does: Provides corrective actions for failure signals such as style drift, role-play/analysis confusion, and straying beyond the public record.

2Core Capabilities

01Dual mode: Trump first-person role-play + third-person behavioral analyst (Analyst by default)
02Behavioral forecasting: Provides probability distributions + confidence ratings + key unknown variables
036 core mental models: Everything is a Transaction, Truthful Hyperbole, Unpredictability as Power, Victimhood Narrative, Zero-Sum Winning, Audience First
048 decision heuristics: Extreme Anchoring, Threat as Leverage, Concession Triggers, Loyalty over Competence, Never Admit Fault, etc.
05Negotiation leverage and zero-sum framework analysis (no win-win, only winners and losers)
06Expressive DNA replication: Ultra-short sentences, absolute vocabulary, nickname system, 'the weave' topic-jumping rhythm
07Counter-example blacklist self-check: Prohibits conceding/showing weakness, prohibits academic jargon, prohibits straying beyond the public record
08Honest boundary self-declaration: Public statements ≠ true intent, unpredictability is genuinely real, domestic political constraints are difficult to track

4Recommended Governance & Evidence

Suggested Controls for Safe Operation
For Agent Runtime
Can be auto-loaded
For Administrators
Register source and version
Evidence & Triggers (File Paths & Excerpts)
  • SKILL.md: Pure prompt instructions and rule text; contains no script execution or external impacts
Audit Notes: Risk Alert: Green, compliant for use. Evaluation target is this skill directory (SKILL.md + references research md files), featuring no scripts, no credential access, no local file writes, and no fixed outbound endpoints; the sole capabilities are prompt-level role-play/behavioral analysis and web search (web access is handled via host model tools). Rated 'pure instruction → green'. Content note: The simulated subject is a political figure; the skill restricts racial/religious inflammatory content within role-play rules to never exceed the scope of the public record. However, methodologies such as 'truthful hyperbole' and 'victimhood narrative' could potentially be misapplied to real-world political communications—this is considered a use-case risk and does not alter the safety rating.

5Second-Pass Verification

  • [ok] 双模式分流机制 — SKILL.md『激活确认(优先执行)』给出触发信号↔模式映射表,角色扮演走路径 A、分析走路径 B、模糊默认 B。
  • [ok] 角色扮演与免责/退出机制 — 有『激活后直接以特朗普身份回应』『首次激活说一次』免责声明与 EXIT TRIGGER 三处原文。
  • [ok] 分析师模式概率化要求 — 『用第三人称,分析特朗普的行为逻辑并给出预判』并注明关键未知变量;反例黑名单第 6 条禁单一确定结论。
  • [ok] 心智模型数量与 description 一致 — 正文『核心心智模型』列模型 1–6,共 6 个,对应 description『6个核心心智模型』。
  • [ok] 决策启发式数量 — 正文『决策启发式』列 1–8 共 8 条,与 description『8条决策启发式』一致。
  • [ok] 安全面复核 — §3 显示凭据/执行/写盘/危险旗标/沙箱均无命中 → security 各数组为空、grade=green。
  • [ok] 元数据与最新动态 — SKILL.md『最新动态(2025-2026)』与资料时间线一致(关税至 145%、IEEPA 部分违宪转 Section 301/232、DOGE 裁减等);schema/skill/meta 保留骨架核定值。

6Verdict & Scenarios

  • Clear dual-mode design: Role-play and analysis are mutually exclusive to prevent output confusion; Analyst mode mandates probabilistic framing for auditable predictions.
  • Comprehensive forecasting framework: 6 mental models + 8 heuristics + concession triggers + 2025-2026 latest updates targeting next moves.
  • Enforces agentic evidence gathering and sets racial/religious content boundaries, providing guardrails for high-risk scenarios.
  • Honest self-boundaries: Explicitly acknowledges that public statements ≠ true intent, that part of the unpredictability is truly random, and that forecasting cannot eliminate uncertainty.
  • Best for: Suitable for analytical scenarios that require decoding the logic behind Trump's public behavior and forecasting his next moves (negotiation, foreign policy, media, appointments, domestic politics); also suitable for role-play/dialogue generation materials.
    Not for: Not suitable for occasions requiring neutrality, politeness, or diplomatic language; not suitable as a methodological textbook for real-world political communication; not suitable for scenarios with extremely high demands on forecast accuracy yet lacking continuous internet tracking capabilities.
    Install Agent-Ready
    ① Verified Mirror 2026-10-10
    Option A · Direct DownloadDownload trump-perspective.tar.gz
    sha256: 6d8f4d82d9c23db4…
    Option B · Copy Guide for Agent
    install.json
    Agents automatically download and verify mirrors from this JSON guide.
    ② Upstream GitHub
    GitHub Repo ↗
    Pinned to commit fe03746870.