Skip to content
Back to search

stock-analysis

This skill should be used when the user asks to analyze this stock, find a target price or buy zone, review fundamentals or technicals, assess risk/earnings/dividends, screen stocks, analyze sector rotation, build an ETF portfolio, design an options strategy, or analyze the macro market; also for Chinese requests such as "分析这只股票", "给出买入区间或目标价", "做基本面/技术面/风险/财报/股息分析", "筛选股票", "分析板块轮动", "构建 ETF 组合", "设计期权策略", "分析宏观市场"; or when they provide a stock ticker, security, ETF, portfolio, earnings event, or investment thesis that requires source-backed market research and an actionable decision.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked yesterday
Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/arthur19940725--stock-analysis-skill/badge.svg)](/listings/arthur19940725--stock-analysis-skill)

HTML

<a href="/listings/arthur19940725--stock-analysis-skill"><img src="/api/listings/arthur19940725--stock-analysis-skill/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

Copy this repository into the Agent's skills directory, for example:

~/.agents/skills/stock-analysis/

Then describe the task in an Agent that supports skills:

Use stock-analysis to analyze Apple Inc. (Nasdaq: AAPL) over a 12-month horizon. Write the report in English.
Use stock-analysis to review the fundamentals and valuation of NVIDIA Corp. (Nasdaq: NVDA). Write the report in English.
Use stock-analysis to assess the technical setup, support levels, and risks of SPDR S&P 500 ETF Trust (NYSE Arca: SPY). Write the report in English.
Use stock-analysis to design a maximum-loss-defined AAPL options strategy using a synchronized bid/ask chain. Write the report in English.

The prompts above demonstrate invocation syntax. They are not frozen datasets or reproducible investment cases.

The skill can also be invoked explicitly:

$stock-analysis Analyze Microsoft Corp. (Nasdaq: MSFT) over a 12-month horizon. Write the report in English.

Before a full analysis, provide the ticker, primary exchange, holding period, benchmark, risk tolerance, and current position status when possible. Missing fields are handled under [`references/research-protocol.md`](references/research-protocol.md) and must be disclosed.

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

SKILL.md · Manifest · 129 lines
  1. ---
  2. name: stock-analysis
  3. description: This skill should be used when the user asks to analyze this stock, find a target price or buy zone, review fundamentals or technicals, assess risk/earnings/dividends, screen stocks, analyze sector rotation, build an ETF portfolio, design an options strategy, or analyze the macro market; also for Chinese requests such as "分析这只股票", "给出买入区间或目标价", "做基本面/技术面/风险/财报/股息分析", "筛选股票", "分析板块轮动", "构建 ETF 组合", "设计期权策略", "分析宏观市场"; or when they provide a stock ticker, security, ETF, portfolio, earnings event, or investment thesis that requires source-backed market research and an actionable decision.
  4. ---
  5. # Stock Analysis
  6. 把股票、ETF、组合、板块、期权与宏观请求转换为可核验、可复算、可证伪的研究结论。围绕一个决策问题组织完整单股研究:
  7. > 在指定持有期内,以当前可验证价格持有该证券,其情景加权总回报是否超过要求回报,并足以补偿永久损失、事件风险和证据不确定性?
  8. ## 先选择运行模式
  9. | 用户意图 | 运行模式 | 加载文件 |
  10. |---|---|---|
  11. | “全面分析”“是否值得买”“目标价”“完整投资报告” | 单股总控 | `references/research-protocol.md`、`references/decision-model.md`、`frameworks/00-thesis-driven-research.md`、`references/output-quality.md` |
  12. | “基本面”“估值”“商业模式”“护城河” | 专项研究 | `references/research-protocol.md` + `frameworks/01-coverage-memo.md` |
  13. | “技术面”“买入区间”“支撑/阻力”“RSI/MACD” | 专项研究 | `references/research-protocol.md` + `frameworks/02-technical-panel.md` |
  14. | “风险”“回撤”“压力测试”“对冲” | 专项研究 | `references/research-protocol.md` + `frameworks/03-risk-framework.md` |
  15. | “财报”“业绩”“EPS”“guidance” | 专项研究 | `references/research-protocol.md` + `frameworks/04-earnings-analyzer.md` |
  16. | “股息”“分红”“收益率”“DRIP” | 专项研究 | `references/research-protocol.md` + `frameworks/05-dividend-analyzer.md` |
  17. | “板块”“轮动”“行业配置” | 专项研究 | `references/research-protocol.md` + `frameworks/06-sector-rotation.md` |
  18. | “筛选”“量化”“多因子”“Top 10” | 专项研究 | `references/research-protocol.md` + `frameworks/07-quant-screener.md` |
  19. | “ETF 组合”“资产配置”“再平衡”“定投” | 专项研究 | `references/research-protocol.md` + `frameworks/08-etf-portfolio.md` |
  20. | “期权”“价差”“covered call”“iron condor” | 专项研究 | `references/research-protocol.md` + `frameworks/09-options-architect.md` |
  21. | “宏观”“利率”“通胀”“美联储”“市场状态” | 专项研究 | `references/research-protocol.md` + `frameworks/10-macro-outlook.md` |
  22. | 财报/公告后更新、持仓复核 | 增量更新 | 原研究结论 + `frameworks/00-thesis-driven-research.md` 的增量流程 |
  23. 命中多个专项时,只加载真正改变决策的文件。完整单股研究不得把多个 framework 并行生成后直接拼接;严格按总控依赖链执行。
  24. ## 输入核验
  25. 先核验证券身份、主交易所、证券类型、报价币种、分析截止时间和持有期。遇到会实质改变结论的歧义时,只问一个最关键问题并停止;否则显式列出默认值后继续。
  26. 至少取得或可靠核验:
  27. - `ticker` 与 `exchange`;
  28. - 当前价格 `P0`、币种、价格时间戳和来源;
  29. - 最近正式财务披露及发布日期;
  30. - 用户任务类型和期望持有期。
  31. 完整单股评级还需要基准、最低要求回报、三情景价值桥、风险压力情景和证据评分。输入字段与默认值见 `references/research-protocol.md`。
  32. ## 取证与数据纪律
  33. 1. 优先读取用户提供的文件、模型与链接;把用户假设与外部事实分开。
  34. 2. 优先监管申报、交易所公告、公司 IR、正式财报、官方宏观数据和产品发行人页面。
  35. 3. 再使用有方法说明的可靠数据源;财经媒体和聚合行情只作补充或价格来源。
  36. 4. 把搜索摘要、论坛和社交媒体仅作为寻找原始资料的线索。
  37. 5. 给每个会改变结论的市场数字附 `as of`;不得把错时价格、共识、利率、IV 或财务数据伪装成同步快照。
  38. 6. 先解析精确标的再调用行情工具。验证市场、证券类型、合约/现货属性与币种,避免同名 ticker 错配。
  39. 7. 若当前环境没有某个特定 MCP,直接使用可用的行情连接器、交易所/发行人页面或可靠网页;不要因固定工具名不存在而停止。
  40. 8. 在可访问来源全部失败后才向用户索取当前价。不得凭记忆估算当前价、估值倍数或期权报价。
  41. 对关键事实使用 `[F]`,外部共识 `[C]`,分析估计 `[E]`,假设 `[A]`,计算推导 `[D]`,无法核验 `[U]`。使用 `K01…` 标识可证伪主张,使用 `S01…` 标识来源。
  42. ## 数据闸门与降级
  43. 缺失数据时降低输出精度,不补造数字:
  44. - 无完整价格序列:不给 RSI、MACD、波动率、beta 或最大回撤,只列所需 OHLCV、参数和观察计划。
  45. - 无点时共识:不用今天的网页数字倒填历史预期,改用公司指引与透明自建估计。
  46. - 无同步双边期权链:只给参数化结构和取链清单,不给具体合约、权利金、Greeks、胜率或“可执行”标签。
  47. - 无历史点时股票池/退市样本/真实发布日期:不声称完成回测,不报 Alpha、Sharpe 或胜率。
  48. - 无资金流数据:不把成交量、持仓或 ETF 价格变化称为机构净流入。
  49. - 无可靠情景概率或价值桥:输出条件估值或暂不评级,不套用固定 20/60/20。
  50. ## 单股总控执行
  51. 加载 `frameworks/00-thesis-driven-research.md`,按以下有向依赖执行:
  52. 1. `M10 → M4`:确定宏观状态、基准、行业利润池与竞争约束。
  53. 2. `M4 → M1 → M8 → M3`:建立经营现实、现金回报与验证日历。
  54. 3. `M2-A → M5`:形成条件估值/反向估值草案,再用因子、盈利修正和价格隐含预期挑战。
  55. 4. `M7 → M2-B`:把风险回写到经营假设、估值参数、情景概率和压力情景,再形成最终目标价。
  56. 5. `M6 + M9 → 总控`:制定现货执行;仅在期权闸门通过时设计期权实施;最后统一计算评级。
  57. 只允许总控层给最终评级。各模块仅输出证据、反证、模型输入和 HANDOFF。若 M6/M9 揭示重大新风险,最多回查一次 `M7 → M2-B`;仍未解决则降低证据分或暂不评级。
  58. ## 确定性计算
  59. 把三情景、要求回报、证据分和否决项写入 JSON,调用:
  60. ```bash
  61. python scripts/decision_math.py path/to/input.json
  62. ```
  63. 正斜杠路径在 Windows 与 Unix 的 Python 上均可使用;Windows 也可用 `path\to\input.json`。
  64. 使用 `scripts/decision_math.py` 计算 `TR_s`、`ETR`、`EU`、`ED`、`PCR`、`SL`、`H`、`ECS` 与机械评级。禁止手工改变阈值以配合叙事。输入结构、公式和评级门槛见 `references/decision-model.md`。
  65. 专项任务若没有完整三情景和证据基础,只给该专项能支持的“技术倾向、风险状态、条件估值、候选清单或参数化结构”,不要冒充完整股票评级。
  66. ## 输出
  67. 默认用简体中文,保留 ticker、指标、机构名和行业术语的英文。短问题先给行动结论,再给依据和失效条件;完整研究使用 `references/output-quality.md` 模板。
  68. 所有行动建议至少包含:
  69. - 数据截止时间与来源层级;
  70. - 当前行动或观察状态;
  71. - 触发/确认条件;
  72. - 技术止损与基本面失效分别列示;
  73. - 目标、风险回报与持有期;
  74. - 数据缺口及其对结论的影响。
  75. 结尾添加:
  76. > 本报告仅供一般信息与研究用途,不构成个性化投资、法律或税务建议。
  77. ## 资源索引
  78. - `references/research-protocol.md`:统一输入、来源、同步、标签、降级与研究预算。
  79. - `references/decision-model.md`:情景回报、证据评分、机械评级和脚本输入。
  80. - `references/output-quality.md`:完整报告模板、检查清单、自动失败与更新纪律。
  81. - `frameworks/00-thesis-driven-research.md`:十模块总控依赖、HANDOFF 和专项工作流。
  82. - 专项框架(与上方路由表一致):
  83. - `frameworks/01-coverage-memo.md`
  84. - `frameworks/02-technical-panel.md`
  85. - `frameworks/03-risk-framework.md`
  86. - `frameworks/04-earnings-analyzer.md`
  87. - `frameworks/05-dividend-analyzer.md`
  88. - `frameworks/06-sector-rotation.md`
  89. - `frameworks/07-quant-screener.md`
  90. - `frameworks/08-etf-portfolio.md`
  91. - `frameworks/09-options-architect.md`
  92. - `frameworks/10-macro-outlook.md`
  93. - `scripts/decision_math.py`:确定性评级计算与输入验证。
  94. - `scripts/test_decision_math.py`:核心边界与降级测试。
README.md · README · 177 lines
  1. # Stock Analysis Skill
  2. English | [中文速览](#中文速览)
  3. ## Overview
  4. A source-first securities research skill for AI agents. It supports stocks, ETFs, sector rotation, earnings, dividends, technical analysis, risk, options, and macro research.
  5. Its purpose is not to produce plausible-sounding investment conclusions. It organizes research into a verifiable, reproducible, and falsifiable decision process.
  6. This README is English-first. The skill currently generates Simplified Chinese by default; add `Write the report in English` to a request when English output is required.
  7. ## Quick start
  8. Copy this repository into the Agent's skills directory, for example:
  9. ```text
  10. ~/.agents/skills/stock-analysis/
  11. ```
  12. Then describe the task in an Agent that supports skills:
  13. ```text
  14. Use stock-analysis to analyze Apple Inc. (Nasdaq: AAPL) over a 12-month horizon. Write the report in English.
  15. Use stock-analysis to review the fundamentals and valuation of NVIDIA Corp. (Nasdaq: NVDA). Write the report in English.
  16. Use stock-analysis to assess the technical setup, support levels, and risks of SPDR S&P 500 ETF Trust (NYSE Arca: SPY). Write the report in English.
  17. Use stock-analysis to design a maximum-loss-defined AAPL options strategy using a synchronized bid/ask chain. Write the report in English.
  18. ```
  19. The prompts above demonstrate invocation syntax. They are not frozen datasets or reproducible investment cases.
  20. The skill can also be invoked explicitly:
  21. ```text
  22. $stock-analysis Analyze Microsoft Corp. (Nasdaq: MSFT) over a 12-month horizon. Write the report in English.
  23. ```
  24. Before a full analysis, provide the ticker, primary exchange, holding period, benchmark, risk tolerance, and current position status when possible. Missing fields are handled under [`references/research-protocol.md`](references/research-protocol.md) and must be disclosed.
  25. ## How it works
  26. The skill first verifies the security identity, price timestamp, currency, and data sources. It then selects the relevant framework. A full single-stock review follows this dependency chain:
  27. ```text
  28. Macro and industry context
  29. -> Operations, financials, and catalysts
  30. -> Conditional valuation and market checks
  31. -> Risk audit and stress testing
  32. -> Execution plan and final rating
  33. ```
  34. Core controls:
  35. - **Source-first**: prefer regulatory filings, exchange notices, company investor relations, and official data.
  36. - **Evidence labels**: separate verified facts `[F]`, consensus `[C]`, estimates `[E]`, assumptions `[A]`, derivations `[D]`, and unknowns `[U]`.
  37. - **Data gates**: reduce precision when data is missing instead of inventing prices, indicators, consensus estimates, or option quotes.
  38. - **Scenario valuation**: calculate expected total return, stress loss, and payoff coverage across bear, base, and bull cases.
  39. - **Deterministic rating**: use [`scripts/decision_math.py`](scripts/decision_math.py) and fixed thresholds instead of changing a rating to fit a narrative.
  40. - **Falsifiable thesis**: attach supporting evidence, the strongest counterevidence, and an invalidation condition to each core claim.
  41. ## Reproducible stock-analysis practice
  42. This practice uses a public issuer, a fixed filing, and explicit verification steps. It is a frozen fundamental-data exercise, not a current recommendation.
  43. ### Research snapshot
  44. | Field | Fixed value |
  45. |---|---|
  46. | Security | Apple Inc. common stock |
  47. | Identifiers | Nasdaq: `AAPL`; SEC CIK `0000320193` |
  48. | Research cutoff | 2025-10-31 16:00 ET (`America/New_York`) |
  49. | Financial period | FY2025, 2024-09-29 through 2025-09-27 |
  50. | Filing | Form 10-K filed 2025-10-31; accession `0000320193-25-000079` |
  51. | Financial units | USD millions, except per-share data |
  52. | Market-price rule | Retrieve the 2025-10-31 regular-session close separately and record its exact field definition |
  53. ### Public sources
  54. | Source ID | Publisher and source | Publication / access date | What to verify |
  55. |---|---|---|---|
  56. | S01 | SEC, [filing detail for accession 0000320193-25-000079](https://www.sec.gov/Archives/edgar/data/320193/000032019325000079/0000320193-25-000079-index.htm) | Published 2025-10-31; accessed 2026-07-31 | Issuer identity, form, filing date, report period, accession, and filing documents |
  57. | S02 | SEC, [Apple FY2025 Form 10-K](https://www.sec.gov/Archives/edgar/data/320193/000032019325000079/aapl-20250927.htm) | Published 2025-10-31; accessed 2026-07-31 | Audited financial statements, product and Services sales, EPS, and cash-flow line items |
  58. | S03 | SEC, [Company Facts JSON](https://data.sec.gov/api/xbrl/companyfacts/CIK0000320193.json) and [EDGAR API documentation](https://www.sec.gov/search-filings/edgar-application-programming-interfaces) | Live API; accessed 2026-07-31 | Programmatic cross-check of standard US-GAAP facts; this is another representation of S02, not independent corroboration |
  59. | S04 | Apple Investor Relations, [historical stock-price lookup](https://investor.apple.com/stock-price/default.aspx) and [FAQ](https://investor.apple.com/investor-relations/faq/default.aspx) | Dynamic; accessed 2026-07-31 | Historical price lookup; Apple states that the quote feed is supplied by Ticker Technologies and delayed by 20 minutes |
  60. | S05 | Nasdaq, [AAPL historical quotes](https://www.nasdaq.com/market-activity/stocks/aapl/historical) | Dynamic; accessed 2026-07-31 | Historical OHLCV and the regular-session price field when the public feed is available |
  61. ### Verified facts and calculations
  62. The following filing values were checked against S02. Standard entity-level facts were also reconciled to S03.
  63. | Metric | FY2025 | FY2024 | Evidence |
  64. |---|---:|---:|---|
  65. | Total net sales | 416,161 | 391,035 | `[F]` S02/S03 |
  66. | Services net sales | 109,158 | 96,169 | `[F]` S02 |
  67. | Net income | 112,010 | 93,736 | `[F]` S02/S03 |
  68. | Diluted EPS (USD/share) | 7.46 | 6.08 | `[F]` S02/S03 |
  69. | Cash generated by operating activities | 111,482 | 118,254 | `[F]` S02/S03 |
  70. | Payments for acquisition of property, plant, and equipment | 12,715 | 9,447 | `[F]` S02/S03 |
  71. Reproducible derivations:
  72. - Total net-sales growth = `416,161 / 391,035 - 1` = **6.43%** `[D]`.
  73. - Services net-sales growth = `109,158 / 96,169 - 1` = **13.51%** `[D]`.
  74. - FY2025 Services mix = `109,158 / 416,161` = **26.23%** `[D]`.
  75. - FY2025 net margin = `112,010 / 416,161` = **26.92%** `[D]`.
  76. - FY2025 free-cash-flow proxy = `111,482 - 12,715` = **USD 98,767 million** `[D]`.
  77. The free-cash-flow proxy is the skill's default calculation of operating cash flow less purchases of property, plant, and equipment. It is not an Apple-reported GAAP line item.
  78. ### How to verify the practice
  79. 1. Open S01 and match `Apple Inc.`, CIK `0000320193`, Form 10-K, report date `2025-09-27`, filing date `2025-10-31`, and accession `0000320193-25-000079`.
  80. 2. In S02, find `Products and Services Performance`, `Consolidated Statements of Operations`, and `Consolidated Statements of Cash Flows`; reconcile the table above before calculating any ratio.
  81. 3. In S03, filter standard XBRL facts by `form=10-K`, `fy=2025`, `fp=FY`, `end=2025-09-27`, and `accn=0000320193-25-000079`. Useful concepts include:
  82. - `us-gaap:RevenueFromContractWithCustomerExcludingAssessedTax`
  83. - `us-gaap:NetIncomeLoss`
  84. - `us-gaap:EarningsPerShareDiluted`
  85. - `us-gaap:NetCashProvidedByUsedInOperatingActivities`
  86. - `us-gaap:PaymentsToAcquirePropertyPlantAndEquipment`
  87. 4. Do not take the last JSON array item without filtering: later filings and comparative periods can repeat the same fact. Verify Services sales in the S02 table because dimensional product data may not appear as an entity-level Company Facts value.
  88. 5. For the historical market price, query 2025-10-31 in S04 and S05. Record:
  89. ```text
  90. ticker | exchange | price field | value | currency | trading date |
  91. timestamp and timezone | delayed status | adjustment definition |
  92. source URL | retrieval time
  93. ```
  94. 6. Use the regular-session close for historical `P0`; do not silently substitute an adjusted close or an intraday/after-hours quote. If only one feed works, disclose that the price lacks a public cross-check. If neither feed works or a discrepancy cannot be resolved, label `P0` as `[U]` and do not issue a price target or final stock rating.
  95. 7. Keep all later information out of the frozen 2025-10-31 analysis to avoid look-ahead bias.
  96. A reproducible prompt:
  97. ```text
  98. Use stock-analysis to review Apple Inc. common stock (Nasdaq: AAPL;
  99. SEC CIK 0000320193) using only information public by 2025-10-31
  100. 16:00 ET. Reconcile the FY2025 facts and calculations in the README
  101. to sources S01-S03. Retrieve the 2025-10-31 regular-session price
  102. from S04 and S05 when available and record its currency, timestamp,
  103. delay, and adjustment definition. Disclose a missing cross-check; if
  104. neither source works or they conflict, mark P0 as [U] and stop before
  105. a target price or final rating. Write the report in English.
  106. ```
  107. The verified snapshot supports a limited operating observation: Services grew faster than total sales in FY2025, while the stated free-cash-flow proxy was USD 98.767 billion. It does not, by itself, supply forecast scenarios, required return, benchmark-relative return, or a defensible buy/hold/sell rating.
  108. ## Repository layout
  109. ```text
  110. stock-analysis/
  111. ├── SKILL.md # Skill entry point and routing
  112. ├── frameworks/ # Research controller and 10 specialist frameworks
  113. ├── references/ # Data protocol, rating model, and output specification
  114. └── scripts/
  115. ├── decision_math.py # Deterministic rating calculation
  116. └── test_decision_math.py # Boundary and degradation tests
  117. ```
  118. Run the tests:
  119. ```bash
  120. python scripts/test_decision_math.py
  121. ```
  122. ## Disclaimer
  123. This project is for general information and research only. It is not personalized investment, legal, or tax advice. Market data is time-sensitive; recheck every conclusion against the latest primary sources before acting.
  124. ## 中文速览
  125. 这是一个面向 AI Agent、以证据和数据闸门为核心的证券研究 skill,覆盖股票、ETF、行业轮动、财报、股息、技术面、风险、期权与宏观市场。README 以英文为主;skill 当前默认输出简体中文,如需英文报告,请在任务中明确写入 `Write the report in English`。
  126. 安装时将仓库复制到 Agent 的 skills 目录,然后提供 ticker、主交易所、持有期、基准、风险承受能力和当前持仓状态。完整规则见 [`SKILL.md`](SKILL.md) 与 [`references/research-protocol.md`](references/research-protocol.md)。
  127. 上面的实践使用 Apple Inc.(Nasdaq: AAPL;CIK `0000320193`)FY2025 Form 10-K 的固定公开数据,并给出 SEC 原文、Company Facts、Apple IR 与 Nasdaq 行情的具体核验路径。财务事实与动态行情分开处理;若历史价格无法公开核验,必须把 `P0` 标为 `[U]`,停止输出目标价和最终评级。
  128. 本项目仅供一般信息与研究用途,不构成个性化投资、法律或税务建议。

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.