← 返回 Skills 市场
harrylabsj

Crypto Exit Rule Drafter

作者 haidong · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ⚠ suspicious
76
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install crypto-exit-rule-drafter
功能描述
A rule-drafting skill that helps users define exit conditions in advance for their crypto positions. Use when the user wants to set pre-defined exit rules. P...
使用说明 (SKILL.md)

crypto-exit-rule-drafter

A rule-drafting skill that helps users define exit conditions in advance for their crypto positions.

Workflow

  1. Ask about the current or planned position: asset, size, entry price, and thesis.
  2. Ask what would make this investment wrong: price drop, fundamental change, or time horizon.
  3. Draft two exit rules: a stop-loss level (downside exit) and a thesis-review level (fundamental exit).
  4. Make the rules concrete and written, not vague intentions.
  5. Set a review trigger and a consequence for when the rule is hit.

Output Format

  • Position summary
  • Downside exit rule (price or percentage level)
  • Fundamental exit rule (what must change)
  • Review trigger and date
  • Written commitment statement

Quality Bar

  • Rules are specific enough to actually trigger, not vague aspirational targets.
  • Separates emotional stopping from rule-based stopping.
  • Does not set stop-losses so tight that normal volatility triggers them.

Edge Cases

  • If the user has no clear thesis for the position, help them articulate one before drafting rules.
  • If the position is long-term and the user has high conviction, focus more on fundamental exit than price stops.

Compatibility

  • Prompt-only, no exchange or portfolio integration.
  • Works from user-provided position details.
安全使用建议
This skill is suspicious because its manifest/README says 'prompt-only' but the shipped handler.py reads a hardcoded, user-specific path (/Users/jianghaidong/.openclaw/skills/...). That could cause the skill to read local files unexpectedly. Before installing or enabling it: (1) ask the author why the handler reads that path and request removal or parameterization of the path (use relative paths or passed-in content instead), (2) review the handler.py yourself or run it in a sandboxed environment to confirm it doesn't access other files or exfiltrate data, (3) ensure the skill only uses data you explicitly supply, and (4) avoid granting it broad filesystem or credential access. If the author cannot justify the local file access or provide a corrected version, do not install it.
功能分析
Type: OpenClaw Skill Name: crypto-exit-rule-drafter Version: 1.0.0 The handler.py file contains a hardcoded absolute path to a specific user's home directory (/Users/jianghaidong/), which is a significant security and portability flaw. Furthermore, the _load_skill_meta function is vulnerable to path traversal because it constructs file paths using the unsanitized skill_name input, allowing potential access to unauthorized directories, although the impact is partially mitigated by the mandatory '/SKILL.md' suffix.
能力标签
crypto
能力评估
Purpose & Capability
SKILL.md describes a prompt-only rule-drafting helper with no integrations or credential needs. The bundled handler.py, however, attempts to load SKILL.md from a hardcoded, user-specific filesystem path (/Users/jianghaidong/.openclaw/skills/{skill_name}/SKILL.md) and loads a prompt template — behavior that is not needed to draft exit rules and is disproportionate to the stated purpose.
Instruction Scope
The runtime instructions (SKILL.md) do not instruct reading any local files or accessing user home directories. The handler code nevertheless opens and reads a local SKILL.md path and returns a template; this is out-of-scope compared to the declared prompt-only workflow and could cause unintended local file reads.
Install Mechanism
There is no install spec (instruction-only), which is lowest-risk for installation. However, the package does include executable code files (handler.py) that will run when the skill is invoked. The absence of an install step reduces disk persistence risk but does not eliminate runtime filesystem access performed by the handler.
Credentials
The skill declares no required env vars or credentials (appropriate for a prompt-only drafter), yet the handler accesses a specific user's home-directory path directly. Requesting or reading arbitrary local files is disproportionate and effectively grants the skill access to local data without any justification in the description.
Persistence & Privilege
always is false and the skill does not declare system-wide changes or persistent installation behavior. Autonomous invocation is allowed (platform default) but does not combine with an 'always' flag or other privilege escalations here.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install crypto-exit-rule-drafter
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /crypto-exit-rule-drafter 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.0.0
Initial release of crypto-exit-rule-drafter - Enables users to define specific, written exit rules for their crypto positions. - Guides users through clarifying position details, risk tolerance, and investment thesis. - Produces two concrete exit rules: price-based (stop-loss) and fundamental-change-based. - Includes a review trigger and commitment statement to encourage disciplined execution. - Designed for prompt-only usage, requiring no portfolio integration.
元数据
Slug crypto-exit-rule-drafter
版本 1.0.0
许可证 MIT-0
累计安装 0
当前安装数 0
历史版本数 1
常见问题

Crypto Exit Rule Drafter 是什么?

A rule-drafting skill that helps users define exit conditions in advance for their crypto positions. Use when the user wants to set pre-defined exit rules. P... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 76 次。

如何安装 Crypto Exit Rule Drafter?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install crypto-exit-rule-drafter」即可一键安装,无需额外配置。

Crypto Exit Rule Drafter 是免费的吗?

是的,Crypto Exit Rule Drafter 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。

Crypto Exit Rule Drafter 支持哪些平台?

Crypto Exit Rule Drafter 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 Crypto Exit Rule Drafter?

由 haidong(@harrylabsj)开发并维护,当前版本 v1.0.0。

💬 留言讨论