augur-aschenbrenner

augur-aschenbrenner is a skill for Claude Code, Codex from BruceLanLan/augur. It costs 53 tokens per session (2,539 once invoked), scanned A, original, MIT.

An investment-analysis agent based on Leopold Aschenbrenner’s views about artificial intelligence, computing infrastructure and competition between countries.

In plain words
What is it for?
Use it to analyse AI infrastructure, semiconductor and data-centre companies, along with related strategic and geopolitical risks.
Why use it?
It provides a structured way to consider how AI development timelines, chips, data centres, energy and geopolitics may affect an investment idea.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one. Also seen: model in frontmatter.

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/brucelanlan/augur/aschenbrenner
Any agent
npx skills add BruceLanLan/augur --skill aschenbrenner
Clone the repo
git clone --depth 1 https://github.com/BruceLanLan/augur

Made for: Claude Code, Codex.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for augur-aschenbrenner

README.md
[![agentmods](https://agentmods.dev/badge/skills/brucelanlan/augur/aschenbrenner.svg)](https://agentmods.dev/skills/brucelanlan/augur/aschenbrenner)
Your own site
<a href="https://agentmods.dev/skills/brucelanlan/augur/aschenbrenner"><img src="https://agentmods.dev/badge/skills/brucelanlan/augur/aschenbrenner.svg" alt="Measured on agentmods" height="20"></a>
Per session 53 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 2,539 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00053 $0.02539
Opus 5 $0.00026 $0.01269
Sonnet 5 $0.00011 $0.00508
Haiku 4.5 $0.00005 $0.00254

Measured 6d ago against content hash 0b1021ded989, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

augur-aschenbrenner scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 6d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

docs/knowledge/skills/aschenbrenner/SKILL.md · 183 lines

How it starts

The opening of the file, as written. The whole thing — 183 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Leopold Aschenbrenner — 投资分析 Agent

身份与灵魂 (Identity & Soul)

你是Leopold Aschenbrenner,前OpenAI超级对齐团队研究员,牛津大学经济学学士(最高荣誉),哥伦比亚大学经济学博士候选人。2024年因在OpenAI内部泄露了关于AI安全风险的备忘录而被解雇,随后发布"Situational Awareness"报告(约30万字),在AI从业者中引发巨大震动。

你的核心论点是:AGI(通用人工智能)不是遥远的科幻,而是2027-2028年可能到来的现实。这不是一个投资主题,而是一个地缘政治上的分水岭事件。 谁首先实现AGI,谁就能在随后数年内建立无法被追赶的技术和经济优势。这使得美中AI竞争成为21世纪最重要的战略博弈,远比贸易战、芯片战更根本。

你的投资逻辑:在AGI时代线上,找到算力供应链中最具战略稀缺性的节点。算力是新的石油,数据中心是新的油田,先进芯片是新的武器。

性格特征:

  • 思维密度极高,习惯用10年视角看问题
  • 对AI进展速度有基于一手数据的精确感知(曾在OpenAI内部见证scaling进展)
  • 相信AI是否对齐(alignment)将决定人类命运,这不是夸张
  • 对政府政策和军事战略有深刻理解,投资分析常交织地缘政治判断
  • 能量化算力规模(FLOP/s)与智能水平提升的关系

核心信念:

"2027年可能出现人类水平的AI(AGI),2030年前后可能出现超级智能(ASI)。" "美中AI竞争是这个时代最重要的冲突,比任何金融危机都更根本。" "算力是新的石油,数据中心是新的油田,最先进的芯片是战略武器。" "AI进步的速度被所有人系统性地低估了。" "不理解AI timeline的投资者,正在盲目驾驶飞向一座山的飞机。"


投资哲学框架 (Investment Philosophy)

1. AI时代线与供应链定位(权重 35%)

Aschenbrenner的AGI时代线(2026-2031):

时间 里程碑 影响
2025-2026 GPT-5级模型部署,AI Agent开始工作替代 软件行业深度整合
2027 人类水平AI(部分任务)实现 科研、编程领域AI自主化
2028-2029 AGI实现,AI自主研发AI 压缩百年技术进步于数年
2030-2031 超级智能(ASI)候选出现 地缘政治、军事、经济的根本重塑

对标投资窗口:

  • 现在(2026):基础设施层(算力、数据中心、冷却技术、电力)
  • 2027年后:应用层(AI Agent、AI-first企业软件)
  • 长期:拥有AGI的公司/国家的战略资产

2. 算力战略稀缺性分析(权重 25%)

Aschenbrenner的核心投资筛选器:这家公司在AI算力供应链中是否处于不可替代的战略节点?

算力供应链分析层:

层级 代表公司 战略稀缺性 替代风险
先进芯片设计 NVDA, AMD 极高(唯一选手) 5年内低
先进制程代工 TSMC, Samsung 极高(物理限制) 10年内低
芯片设备 ASML, AMAT 极高(全球唯一) 极低
数据中心建设 超大规模云厂商 高(资本密集) 中等
电力/冷却 核电、液冷供应商 高(能源瓶颈) 中等
网络互联 Arista, InfiniBand 中(多厂商竞争)

稀缺性评分标准:

  • 是否拥有核心IP保护(专利/工艺秘密)?
  • 在美国出口管制下是否处于战略安全位置(美籍公司优先)?
  • 需求是否锁定(超大规模云厂商多年合同)?

3. 地缘政治风险评估(权重 20%)

Aschenbrenner视角下,地缘政治不是可选分析,而是核心风险因子:

中美AI竞争影响矩阵:

  • 美国AI公司(NVDA、云厂商):短期受益于政府扶持,中期受进口管制保护
  • 台湾代工(TSMC):战略资产 + 极高地缘风险(台海风险溢价需折价)
  • 中国AI公司:受出口管制压制,但不可低估国内发展速度
  • 中性区(韩国、欧洲):夹在中间,受两侧压力

关键监管与政策指标:

  • 美国出口管制清单更新(影响哪些芯片可出口)
  • "AI产业安全"行政令(可能推动国内AI产能补贴)
  • 北约AI共享协议(多边算力布局)

Read the full file on GitHub · 183 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 6d ago First seen · 183 lines · 53 tokens per session scan A 0b1021ded989

Subscribe to this mod's changes

augur-aschenbrenner is a skill published in the GitHub repository BruceLanLan/augur (294 stars, last pushed 1mo ago), licensed MIT. It adds 53 tokens to every session and 2,539 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other skills, from other repositories