论文审计
- 作者仓库星标 640
- 许可证 MIT
- 作者更新于 实时读取
- 作者仓库 honeclaw
- 领域
- 通用
- 兼容 Agent
-
- Claude Code
- Cursor
- Cline
- Codex
- Windsurf
- Gemini CLI
- +20
- 信任分
- 94 / 100 · 已通过审计
- 作者 / 版本 / 许可
- @B-M-Capital-Research · MIT
- Token 消耗评级
- 低消耗
- 接入复杂程度
- 即装即用
- 是否需要外部 API Key
- 不需要
- 兼容的系统
- 未声明（默认跨平台）
- 底层运行要求
- 无特殊要求
- 文件与系统权限
-
- 只读
- 允许写入 / 修改
- 网络行为
- 允许外网请求
- 安装命令数
- 26 条
档案由构建时根据 SKILL.md 与安装命令自动衍生，可能与作者实际意图存在差异。
需要注意： 未限定 allowed-tools，默认拥有全部工具权限。
---
name: Stock Research
description: Canonical Hone equity-research skill covering single-stock analysis, valuation framing, and crit…
category: 通用
runtime: 无特殊运行时
---
# Stock Research 输出预览
## PART A: 任务判断
- 适用问题：通用任务拆解、检查和交付。
- 输入要求：目标材料、限制条件、期望输出和验收方式。
- 证据边界：围绕“Stock Research Skill / Tool Guide / Mode Selection”读取原文规则，不把推断写成作者承诺。
## PART B: 执行结果
- **01** 任务判断：确认你的需求是否属于通用任务拆解、检查和交付，并标出输入、限制和预期结果。
- **02** 执行计划：优先按“Stock Research Skill / Tool Guide / Mode Selection”拆成步骤，说明每一步会读取什么、修改什么、产出什么。
- **03** 交付结果：给出可复制的命令、文件改动、检查清单或内容草稿，并说明如何继续迭代。
- **04** 风险边界：结合 读取文件、写入/修改文件、会按任务需要访问外部网络、通常不需要额外 API Key 给出执行前确认项。
## Running Rules
- 读取文件、写入/修改文件；会按任务需要访问外部网络；通常不需要额外 API Key。
- 先小样例验证，再放大到真实任务。
- 交付时同时给结果、检查口径和下一步迭代建议。 先确认触发方式
原文没有稳定的斜杠命令要求。安装验证后通常全局生效，直接在对话里点名这个 Skill 并描述任务即可。
给清楚输入和边界
告诉 Agent 目标文件或材料、期望结果、不可改范围、是否允许联网或执行命令。本 Skill 的权限画像是：读取文件、写入/修改文件。
小样例验证后再放大
先用一个小任务确认它会围绕“Stock Research Skill / Tool Guide / Mode Selection”工作；涉及文件或命令时，先看 diff、日志、预览或测试结果。
复核后再交付
检查最终产物是否包含明确结果、必要证据和下一步动作；如果输出泛泛而谈，就补充输入、边界和验收标准后重跑。
---
name: Stock Research
description: Canonical Hone equity-research skill covering single-stock analysis, valuation framing, and crit…
category: 通用
source: B-M-Capital-Research/honeclaw
---
# Stock Research
## 什么时候使用
- 把通用方向的常用动作沉淀成 Agent 可调用的技能 适合处理通用任务拆解、检查、交付和复盘，核心价值是把输入、判断、执行、验证和交付边界固定下来，避免 Agent 泛泛回答。 把任务拆成可执行、可检查、可继续迭代的步骤；通常不需要额外…
- 面向通用任务拆解、检查和交付，优先处理能明确输入、步骤和验收标准的工作。
## 需要提供什么
- 目标材料、目录范围、期望结果和不可改动内容。
- 是否允许联网、执行命令、读写文件或调用外部服务。
## 执行规则
- 围绕「Stock Research Skill / Tool Guide / Mode Selection」组织步骤，不把推断写成作者事实。
- 读取文件、写入/修改文件；会按任务需要访问外部网络；通常不需要额外 API Key。
- 先跑小样例，确认结果可检查后再扩大任务范围。
## 输出要求
- 给出最终产物、关键证据、验证方式和下一步动作。
- 信息不足时标记 unknown，不编造命令、平台或依赖。 证据边界与执行链路
作者原文负责流程事实；仓库文件负责来源和命令；流狐只补充适用场景、限制和质量判断。
skill "Stock Research" {
输入层 -> 用户目标 + 目标文件 + 禁止范围 + 验收标准
上下文层 -> Stock Research Skill / Tool Guide / Mode Selection
规则层 -> SKILL.md 触发条件 / 执行顺序 / 输出格式
运行层 -> 无特殊运行时 | 读取文件、写入/修改文件 | 会按任务需要访问外部网络
安全层 -> 通常不需要额外 API Key + 小任务验证 + diff / 日志复核
输出层 -> 可复制结果 + 检查清单 + 下一步迭代
}Stock Research Skill
This is the canonical equity-research entrypoint for Hone.
Use it for three closely related user intents:
- Single-company research
- Valuation framing for a named company
- Criteria-based stock screening that returns a short comparison list
Prefer keeping these modes inside one skill so the model does not have to choose between overlapping prompt variants.
Tool Guide
| Tool call | Purpose |
|---|---|
data_fetch(data_type="snapshot", symbol="ticker") |
Recommended. Fetch a snapshot with price action plus company overview |
data_fetch(data_type="quote", symbol="ticker") |
Fetch detailed real-time quote data such as price, change, and volume |
data_fetch(data_type="profile", symbol="ticker") |
Fetch company details such as business description, industry, and CEO |
data_fetch(data_type="financials", symbol="ticker") |
Fetch financial statements or valuation-relevant fundamentals |
data_fetch(data_type="gainers_losers") |
Broader market scan when a screening request needs candidates |
data_fetch(data_type="sector_performance") |
Sector strength context for screening or relative positioning |
web_search(query="...") |
Search for news, analyst views, and recent events |
Mode Selection
Choose the mode from the user's request before fetching data:
- Research mode: the user asks about one company, ticker, fundamentals, technicals, or recent developments
- Valuation mode: the user asks whether a company looks rich, cheap, stretched, fairly priced, or wants a valuation bridge / peer view
- Screening mode: the user asks for a shortlist that matches factors such as AI, dividend yield, value, growth, or momentum
Research Mode
- Identify the ticker mentioned by the user. If it is unclear, search first with
data_fetch(data_type="search", symbol="...") - Call
snapshotfor the baseline data - Decide whether to add
web_searchfor news or causes - Output a combined answer covering price action, fundamentals, recent events, and risks
- If the user explicitly asks for a chart, trend line, comparison visual, or the answer would be materially clearer as a chart, hand off to
chart_visualizationwith the concrete numbers you already fetched
Valuation Mode
- Resolve the ticker first; do not attempt valuation without confirming the company
- Fetch
financials; addquoteorsnapshotif you also need current market context - Use
web_searchfor the latest operating updates, guidance changes, or peer-comparison context - Explain the valuation through assumptions, peer multiples, and business quality, and state which conditions would make the company look richer, more balanced, or more compelling relative to peers
- Do not collapse the result into a simplistic categorical verdict with no assumptions attached
Screening Mode
- Extract the user's explicit criteria before naming companies
- Use
gainers_losers,sector_performance, or targetedweb_searchto form an initial candidate set - Narrow the result to 3-5 names and fetch
snapshotfor each final candidate - Return a comparison shortlist with why each name matches the screen, plus the main risk or diligence gap for each one
- Do not output a blunt recommendation list without comparison logic or caveats
先判断是否适合
作者设计意图
作者的方法与取舍
边界和复核