基于 SOC 职业分类
用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt;运行前请先检查来源。
npx skills add https://github.com/lev-os/agents --skill circle-of-competence命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地?可下载 SkillsMP 当前能够提供的文件。
正在显示 SKILL.md
| name | circle-of-competence |
| description | Define and operate within your areas of genuine expertise to avoid costly mistakes from overconfidence |
Circle of Competence is Warren Buffett and Charlie Munger's mental model for limiting decisions to areas where you have genuine understanding and expertise. The core insight: knowing the boundaries of your competence is more important than the size of that competence. Operating outside your circle leads to predictable failures because confidence diminishes slower than competence as you cross the perimeter.
List domains where you have genuine expertise - not interest, but deep understanding. Ask: Can you explain how this business/domain makes money? Can you predict second-order effects? Can you identify what separates good from great performers?
Example: An enterprise SaaS engineer might list: cloud infrastructure, API design, database optimization, but NOT consumer mobile apps or hardware.
For every opportunity, categorize it: "In" (within circle), "Out" (clearly outside), or "Too Hard" (uncertain). Charlie Munger: "You don't have to be a jack-of-all-trades." Most opportunities should land in "Out" or "Too Hard" - that's healthy selectivity.
Before committing, ask: Can I read this company's earnings report and form independent questions? Can I isolate where this sits among competitors? Can I explain edge cases and failure modes? If you're guessing or parroting others, you're outside your circle.
As you approach your circle's edge, confidence doesn't drop as fast as competence - this is the danger zone. Watch for rationalization ("I can figure this out"), unfounded optimism, or reliance on others' expertise. These signal you're crossing the boundary.
When facing an attractive opportunity outside your circle, resist the urge to expand. Buffett: "If we can't find things within our circle of competence, we won't expand the circle. We'll wait." Deliberate expansion takes years; hasty expansion invites disaster.
Only expand your circle through sustained, focused study - not because an opportunity appeared. Spend years reading, practicing, and learning from mistakes in a new domain before considering it "inside" your circle.
Situation: Buffett was approached about investing in tech companies during the dot-com boom (1999-2000). Everyone was making fortunes; pressure to participate was immense.
Application: He categorized technology as "Too Hard" - he couldn't predict which companies would win or explain their competitive moats. Despite ridicule for "missing out," he stuck to businesses he understood: insurance, banking, consumer brands.
Outcome: He avoided $5 trillion in losses when the bubble burst (2000-2002), while his "boring" holdings remained stable. His discipline to say "I don't know" preserved billions in capital.
Routes adaptive multi-agent deliberation with fractal context cycles. Use when using /cdo, think/deep/debug/parliament work, or long runs paired with autoresearch scheduling.
Use when you need an evidence-first provenance report from `cass`, repo docs, `.lev`, `~/.agents/diagrams`, `qmd`, or Grep/Glob tools.
Use the codebase knowledge graph for structural code queries. Triggers on: explore the codebase, understand the architecture, what functions exist, show me the structure, who calls this function, what does X call, trace the call chain, find callers of, show dependencies, impact analysis, dead code, unused functions, high fan-out, refactor candidates, code quality audit, graph query syntax, Cypher query examples, edge types, how to use search_graph.