When to Use
Use this skill when a user is evaluating real estate investing decisions such as buy-and-hold rentals, value-add deals, BRRRR projects, flips, or house hacks.
Agent handles strategy selection, return targets, deal triage, underwriting, financing pressure tests, diligence sequencing, portfolio fit, and exit discipline.
Architecture
Memory lives in ~/real-estate-investing/. If ~/real-estate-investing/ does not exist, run setup.md. See memory-template.md for the baseline structures.
CODEBLOCK0
Quick Reference
| Topic | File | Use it for |
|---|
| First-run activation | INLINECODE4 | Integration behavior, boundaries, and storage scope |
| Memory baseline |
memory-template.md | Create local files for strategy, pipeline, and decisions |
| Strategy and buy box |
thesis-and-box.md | Define goals, market focus, and non-negotiable filters |
| Strategy fit and return targets |
strategy-selection.md | Match investing styles to capital, time, and risk |
| 30-second screening |
deal-triage.md | Reject weak deals before deep work |
| Underwriting model |
underwriting.md | Revenue, expenses, reserves, and scenario stress tests |
| Debt and downside |
financing-and-risk.md | Loan structure, DSCR, reserves, and refinance risk |
| Diligence workflow |
diligence-and-red-flags.md | Documents to request and kill-shot risks |
| Operations and exits |
portfolio-ops.md | Management, capex planning, and sell-or-hold triggers |
Requirements
- - No credentials or external services required by default.
- Ask before storing exact addresses, personal legal names, lender documents, or tax IDs.
- Treat legal, tax, insurance, and lending guidance as decision support, not licensed advice.
Investor Lane
This skill is for investment decisions, not broad real-estate generalism.
- - Use this skill for return-driven property choices, capital allocation, rentability, strategy fit, leverage risk, and portfolio growth decisions.
- Do not use this skill as the primary router for generic buyer, seller, agent, or landlord workflows that are not investment-led.
- If the user mainly needs transaction help outside investing, route to
real-estate-skill, home-buying, or rental as appropriate.
Deal Loop
Use this order unless the user explicitly wants a narrower question answered:
- 1. Lock thesis and buy box.
- Match the strategy to capital, time, and operational reality.
- Run 30-second triage.
- Underwrite with conservative assumptions.
- Stress debt, timeline, vacancy, and capex.
- Sequence diligence around the biggest ways the deal can die.
- Check operational fit and exit routes before recommending action.
Core Rules
1. Start With the Thesis, Not the Listing
- - Define target strategy, hold period, capital budget, target return, and acceptable workload before reviewing deals.
- A good-looking property outside the thesis is usually a distraction, not an opportunity.
2. Choose the Right Strategy Before Chasing Yield
- - Match the deal shape to the real strategy: buy-and-hold, value-add, BRRRR, flip, house hack, or short-term rental.
- A strategy with higher headline returns is worse if it exceeds the user's time, execution skill, or capital resilience.
3. Kill Weak Deals Early
- - Use
deal-triage.md before any full underwriting pass. - Reject deals fast when rent reality, neighborhood fit, repair scope, financing feasibility, or execution complexity already breaks the plan.
4. Underwrite to Reality, Not Broker Story
- - Use in-place rent, market rent, vacancy, repairs, management, taxes, insurance, utilities, turnover, and capital reserves explicitly.
- Every optimistic assumption needs a downside case beside it.
5. Read the Full Return Stack
- - Judge rentability and profitability through NOI, cap rate, cash-on-cash, DSCR, break-even occupancy, equity creation, and exit optionality together.
- A deal is not "profitable" just because one metric looks great.
6. Stress the Debt Before Trusting the Return
- - Test break-even occupancy, DSCR, rate shocks, delayed refinance, slower lease-up, and higher rehab cost.
- If the deal only works under easy debt, it does not really work.
7. Match Strategy to Operations
- - Favor strategies the user can actually operate: tenant quality, contractor depth, management bandwidth, market distance, and legal complexity matter as much as purchase price.
- A lower-return deal with cleaner operations can be better than a "high ROI" deal that will fail in execution.
8. Treat Diligence as Risk Pricing
- - Convert every unknown into one of four outcomes: verify, renegotiate, reserve for it, or walk.
- Never hand-wave title issues, insurance friction, deferred maintenance, or rent-roll quality.
9. Decide With a Written Kill-Switch
- - Before recommending "buy," state what would make the answer become "no."
- Store the reason for each passed or rejected deal so future judgments improve instead of repeating the same mistake.
Real-Estate-Investing Traps
| Trap | Why It Fails | Better Move |
|---|
| Using pro forma rents as fact | Return targets look safe when they are not | Anchor to proven in-place rent and verified market comps |
| Underestimating capex |
Cash flow looks healthy until the first major repair | Separate maintenance, turnover, and true capital expense reserves |
| Calling any positive cash flow "rentable" | Thin margins hide vacancy, delinquency, and management drag | Test DSCR, break-even occupancy, and reserves before calling a rental healthy |
| Treating cash-on-cash as the only metric | Leverage can fake a great return on a weak asset | Read DSCR, break-even occupancy, debt terms, and exit risk together |
| Buying outside operational competence | Distance and vendor weakness destroy execution | Match asset type and market to actual operating capacity |
| Assuming refinance is guaranteed | BRRRR math breaks when rates, value, or DSCR move | Model delayed or failed refinance before acquisition |
| Ignoring taxes, insurance, and regulation drift | "Stable" deals can reprice overnight | Track reassessment risk, insurance availability, and local restrictions |
| Falling in love with one deal | Emotion overrides guardrails and exceptions stack up | Compare every deal against the thesis and recent rejects |
Data Storage
Local state lives in ~/real-estate-investing/:
- - the local strategy memory file for buy box and recurring guardrails
- the local pipeline file for active deals and blockers
- the local market-notes file for rents and recurring local risks
- the local decisions log for won/lost deals and post-mortems
- INLINECODE18 for inactive opportunities and old decision logs
Security & Privacy
Data that stays local:
- - strategy preferences, underwriting assumptions, market notes, and decision logs in INLINECODE19
Data that leaves your machine:
This skill does NOT:
- - place offers, sign contracts, or move money
- claim jurisdiction-specific legal or tax certainty without source material
- submit lender or insurance applications automatically
- store secrets, bank logins, or full identity packages
- modify its own INLINECODE20
Scope
This skill ONLY:
- - helps choose the right investing strategy before analyzing properties
- helps analyze and compare real estate investment opportunities
- keeps local memory about strategy, markets, and deal outcomes
- improves decisions through explicit assumptions, stress tests, and post-mortems
This skill NEVER:
- - act as a generic buyer, seller, or agent assistant
- guarantee returns
- replace licensed legal, tax, insurance, or lending advice
- recommend a deal without stating key assumptions and failure points
Related Skills
Install with
clawhub install <slug> if user confirms:
- -
property-valuation - Estimate market value and compare comp-driven pricing against the investment thesis. - INLINECODE23 - Separate owner-occupant decisions from pure investment decisions when a deal mixes both.
- INLINECODE24 - Evaluate tenant-landlord realities, leasing friction, and rent-side execution details.
- INLINECODE25 - Compare property returns against broader investing trade-offs and capital allocation choices.
- INLINECODE26 - Broader transaction support across buying, selling, agents, and mixed property roles.
Feedback
- - If useful: INLINECODE27
- Stay updated: INLINECODE28
何时使用
当用户正在评估房地产投资决策时使用此技能,例如买入并持有出租、增值交易、BRRRR项目、翻新转售或房屋黑客策略。
智能体负责处理策略选择、回报目标、交易筛选、承销评估、融资压力测试、尽职调查排序、投资组合适配以及退出纪律。
架构
记忆文件存储在 ~/real-estate-investing/ 目录下。如果 ~/real-estate-investing/ 不存在,请运行 setup.md。基础结构请参见 memory-template.md。
text
~/real-estate-investing/
├── memory.md # 策略、护栏和当前优先级
├── pipeline.md # 正在审核的交易及其当前阶段和障碍
├── markets.md # 市场笔记、租金假设和本地风险
├── decisions.md # 已赢/已失交易、事后复盘和模式更新
└── archive/ # 非活跃想法和旧交易记录
快速参考
| 主题 | 文件 | 用途 |
|---|
| 首次运行激活 | setup.md | 集成行为、边界和存储范围 |
| 记忆基线 |
memory-template.md | 为策略、交易管道和决策创建本地文件 |
| 策略和买入框 | thesis-and-box.md | 定义目标、市场焦点和不可协商的筛选条件 |
| 策略适配和回报目标 | strategy-selection.md | 将投资风格与资本、时间和风险相匹配 |
| 30秒筛选 | deal-triage.md | 在深入工作前拒绝弱势交易 |
| 承销模型 | underwriting.md | 收入、支出、储备和情景压力测试 |
| 债务和下行风险 | financing-and-risk.md | 贷款结构、DSCR、储备和再融资风险 |
| 尽职调查流程 | diligence-and-red-flags.md | 需要索取的文件和致命风险 |
| 运营和退出 | portfolio-ops.md | 管理、资本支出规划和卖出或持有触发条件 |
要求
- - 默认情况下不需要任何凭证或外部服务。
- 在存储确切地址、个人法定姓名、贷款文件或税号之前需先询问。
- 将法律、税务、保险和贷款指导视为决策支持,而非持牌建议。
投资者赛道
此技能专为投资决策设计,而非广泛的房地产通识。
- - 使用此技能进行以回报为导向的房产选择、资本配置、可租性、策略适配、杠杆风险和投资组合增长决策。
- 请勿将此技能作为通用买方、卖方、经纪人或房东工作流程的主要路由,除非这些流程以投资为主导。
- 如果用户主要需要投资以外的交易帮助,请相应路由至 real-estate-skill、home-buying 或 rental。
交易循环
除非用户明确要求回答更窄范围的问题,否则按此顺序执行:
- 1. 锁定投资主题和买入框。
- 将策略与资本、时间和运营现实相匹配。
- 运行30秒筛选。
- 使用保守假设进行承销。
- 对债务、时间线、空置率和资本支出进行压力测试。
- 围绕交易可能失败的最大风险点排序尽职调查。
- 在推荐行动前检查运营适配性和退出路径。
核心规则
1. 从主题开始,而非房源
- - 在审查交易前,先定义目标策略、持有期、资本预算、目标回报和可接受的工作量。
- 主题之外的好看房产通常是干扰项,而非机会。
2. 在追逐收益前选择正确的策略
- - 将交易形态与真实策略相匹配:买入并持有、增值、BRRRR、翻新转售、房屋黑客或短期租赁。
- 如果一项策略超出用户的时间、执行技能或资本韧性,即使名义回报更高,也是更差的选择。
3. 尽早淘汰弱势交易
- - 在任何完整的承销评估前使用 deal-triage.md。
- 当租金现实、社区适配度、维修范围、融资可行性或执行复杂度已经破坏计划时,快速拒绝交易。
4. 基于现实承销,而非经纪人的故事
- - 明确使用现有租金、市场租金、空置率、维修、管理、税费、保险、水电费、租户周转和资本储备。
- 每个乐观假设旁边都需要一个下行情景。
5. 阅读完整的回报堆栈
- - 通过NOI、资本化率、现金回报率、DSCR、盈亏平衡入住率、权益创造和退出可选性综合判断可租性和盈利能力。
- 仅仅因为一个指标看起来不错,交易并不算盈利。
6. 在信任回报前先对债务进行压力测试
- - 测试盈亏平衡入住率、DSCR、利率冲击、延迟再融资、较慢的租赁去化和更高的翻新成本。
- 如果交易仅在宽松债务条件下可行,则实际上并不可行。
7. 将策略与运营相匹配
- - 优先选择用户实际能够运营的策略:租户质量、承包商深度、管理带宽、市场距离和法律复杂度与购买价格同等重要。
- 运营更干净的较低回报交易可能优于执行中会失败的高ROI交易。
8. 将尽职调查视为风险定价
- - 将每个未知转化为四种结果之一:核实、重新谈判、为此预留储备或退出。
- 切勿对产权问题、保险摩擦、递延维护或租金册质量含糊其辞。
9. 用书面的终止开关做决策
- - 在推荐买入之前,说明什么情况会使答案变为否。
- 存储每个通过或拒绝交易的原因,以便未来判断能够改进,而非重复同样的错误。
房地产投资陷阱
| 陷阱 | 失败原因 | 更好的做法 |
|---|
| 将预估租金视为事实 | 回报目标看似安全,实则不然 | 锚定已验证的现有租金和经核实的市场可比数据 |
| 低估资本支出 |
现金流看似健康,直到第一次大修 | 将维护、周转和真实资本支出储备分开 |
| 将任何正现金流称为可租 | 微薄利润掩盖了空置、拖欠和管理拖累 | 在判定出租健康前测试DSCR、盈亏平衡入住率和储备 |
| 将现金回报率视为唯一指标 | 杠杆可以在弱势资产上伪造出高回报 | 综合阅读DSCR、盈亏平衡入住率、债务条款和退出风险 |
| 在运营能力范围外购买 | 距离和供应商弱点会破坏执行 | 将资产类型和市场与实际运营能力相匹配 |
| 假设再融资有保障 | BRRRR计算在利率、价值或DSCR变动时崩溃 | 在收购前模拟延迟或失败的再融资 |
| 忽视税收、保险和监管变动 | 稳定的交易可能一夜之间重新定价 | 追踪重估风险、保险可用性和本地限制 |
| 爱上某一笔交易 | 情绪覆盖了护栏,例外不断累积 | 将每笔交易与主题和近期拒绝的交易进行比较 |
数据存储
本地状态存储在 ~/real-estate-investing/ 中:
- - 用于买入框和重复性护栏的本地策略记忆文件
- 用于活跃交易和障碍的本地交易管道文件
- 用于租金和重复性本地风险的市场笔记文件
- 用于已赢/已失交易和事后复盘的本地决策日志
- archive/ 用于非活跃机会和旧决策日志
安全与隐私
留在本地的数据:
- - 策略偏好、承销假设、市场笔记和决策日志存储在 ~/real-estate-investing/ 中
离开您机器的数据:
此技能不会:
- - 提交报价、签署合同或转移资金
- 在没有原始资料的情况下声称具有特定司法管辖区的法律或税务确定性
- 自动提交贷款或保险申请
- 存储密码、银行登录信息或完整身份信息包
- 修改自身的 SKILL.md
范围
此技能仅:
- - 在分析房产前帮助选择正确的投资策略
- 帮助分析和比较房地产投资机会
- 在本地保存关于策略、市场和交易结果的记忆
- 通过明确的假设、压力测试和事后复盘改进决策
此技能从不:
- - 充当通用买方、卖方或经纪人助手
- 保证回报
- 替代持牌法律、税务、保险或贷款建议
- 在未说明关键假设和失败点的情况下推荐交易
相关技能
在用户确认后使用 clawhub install
安装:
- - property-valuation - 估算市场价值,并将可比驱动定价与投资主题进行比较。
- home-buying - 当交易混合了自住和纯投资决策时,将自住决策与纯投资决策分开。
- rental - 评估租户-房东现实、租赁摩擦和租金侧执行细节。
- invest - 将房产回报与更广泛的投资权衡和资本配置选择进行比较。
- real-estate-skill -