Agent助手
- 作者仓库星标 6,018
- 作者更新于 实时读取
- 作者仓库 OctoBot
- 领域
- AI 智能
- 兼容 Agent
-
- Claude Code
- Cursor
- Cline
- Codex
- Windsurf
- Gemini CLI
- +20
- 信任分
- 88 / 100 · 社区维护
- 作者 / 版本 / 许可
- @Drakkar-Software · 未声明 license
- Token 消耗评级
- 低消耗
- 接入复杂程度
- 需简单配置
- 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
- Shell 执行
- 允许写入 / 修改
- 网络行为
- 仅限本地
- 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: portfolio-distribution
description: Use this skill for making final portfolio allocation decisions, determining optimal position siz…
category: AI 智能
runtime: 无特殊运行时
---
# portfolio-distribution 输出预览
## PART A: 任务判断
- 适用问题：提示词、Agent 工作流、模型评估或自动化推理。
- 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“Overview / Instructions / 1. Position Sizing Framework”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提示词、Agent 工作流、模型评估或自动化推理，并标出输入、限制和预期结果。
- **02** 执行计划：优先按“Overview / Instructions / 1. Position Sizing Framework”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、执行终端命令、写入/修改文件、主要在本地完成、通常不需要额外 API Key 给出执行前确认项。
## Running Rules
- 读取文件、执行终端命令、写入/修改文件；主要在本地完成；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
原文出现了 `/memories` 这类斜杠命令；如果你的 Agent 支持命令触发，优先用命令开场，再补充目标和边界。
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、执行终端命令、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“Overview / Instructions / 1. Position Sizing Framework”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: portfolio-distribution
description: Use this skill for making final portfolio allocation decisions, determining optimal position siz…
category: AI 智能
source: Drakkar-Software/OctoBot
---
# portfolio-distribution
## 什么时候使用
- 把AI / Agent方向的常用动作沉淀成 Agent 可调用的技能 适合处理AI Agent、提示词、模型评估与自动化推理，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…
- 面向提示词、Agent 工作流、模型评估或自动化推理，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「Overview / Instructions / 1. Position Sizing Framework」组织步骤，不把推断写成作者事实。
- 读取文件、执行终端命令、写入/修改文件；主要在本地完成；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 "portfolio-distribution" {
输入层 -> 用户目标 + 目标文件 + 禁止范围 + 验收标准
上下文层 -> Overview / Instructions / 1. Position Sizing Framework
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、执行终端命令、写入/修改文件 | 主要在本地完成
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
}portfolio-distribution
Overview
This skill provides guidance on translating risk assessments and market analyses into specific portfolio allocation decisions. Your role is to determine optimal position sizes while respecting risk management constraints.
Instructions
1. Position Sizing Framework
Base Position Size Calculation
Base_Size = Portfolio_Risk_Percentage × (Account_Size / Risk_Per_Trade)
Adjusted_Size = Base_Size × Confidence_Modifier × Risk_ModifierConfidence Modifiers
- Very high confidence (>0.85): 1.5x
- High confidence (0.70-0.85): 1.0x
- Medium confidence (0.50-0.70): 0.75x
- Low confidence (<0.50): 0.25x or skip
Risk Modifiers (from Risk Judge)
- Low overall risk: 1.2x
- Medium overall risk: 1.0x
- High overall risk: 0.5x
2. Portfolio Risk Management
Total Exposure Limits
- Maximum total position size: 80% of portfolio
- Single position maximum: 25% of portfolio
- Correlated positions combined: <40% of portfolio
- Reserve minimum: 20% cash for opportunities
Diversification Rules
- Don't concentrate >40% in one sector
- Consider correlation between positions
- Balance long and short exposure when appropriate
- Maintain liquidity for rebalancing
3. Action Determination
Buy Actions
- Bull case won with high confidence
- Current position < target allocation
- Entry zone identified with favorable risk/reward
- Liquidity adequate for execution
Sell Actions
- Bear case won with high confidence
- Current position > target allocation
- Stop loss hit or target reached
- Risk/reward deteriorated
Hold Actions
- Unclear winner or low confidence
- Current allocation within target range
- Waiting for better entry/exit
- Preserving capital for better opportunities
4. Execution Planning
Priority Ordering
- Highest conviction trades first
- Most time-sensitive opportunities
- Rebalancing underweight positions
- Taking profits on overweight positions
Timing Recommendations
- Immediate: High confidence, favorable conditions
- Limit Order: Medium confidence, specific entry target
- Wait: Low confidence or better setup expected
- Scale In/Out: Uncertainty about timing, reduce entry/exit risk
5. Trade Structuring
Entry Strategy
- Single entry: High confidence, clear signal
- Scaled entry: Medium confidence, want better average
- Layered limit orders: Range-bound market
- Market order: Time-sensitive, high urgency
Exit Strategy
- Initial stop loss: Below key support (long) or above resistance (short)
- First target: 1.5-2R (risk units)
- Second target: Major resistance/support
- Final: Trailing stop to ride trend
Output Format
{
"distribution": {
"allocations": [
{
"symbol": "BTC/USDT",
"action": "buy",
"current_percentage": 10.0,
"target_percentage": 18.0,
"quantity_change": 0.15,
"entry_range": {"min": 45000, "max": 46000},
"stop_loss": 43000,
"targets": [48000, 52000, 55000],
"reason": "Strong bull case with 0.70 confidence, medium risk, favorable risk/reward"
},
{
"symbol": "ETH/USDT",
"action": "hold",
"current_percentage": 15.0,
"target_percentage": 15.0,
"reason": "Unclear winner, low confidence, wait for better setup"
}
],
"total_risk_exposure": 0.65,
"cash_reserve": 0.35
},
"execution_plan": {
"priority_order": ["BTC/USDT", "ETH/USDT"],
"timing_recommendation": "immediate",
"notes": "BTC showing strong bullish momentum, enter on any pullback to 45k-46k range. Set stops below 43k. Scale out at targets."
},
"risk_summary": {
"portfolio_risk": "medium",
"max_drawdown_estimate": "12%",
"correlation_warning": "BTC and ETH highly correlated, don't oversize both"
}
}Best Practices
- Never overleverage - Respect maximum position sizes
- Always define stops - Know your exit before entry
- Scale positions by conviction - Higher confidence = larger size
- Consider correlation - Don't concentrate risk
- Maintain reserves - Keep cash for opportunities
- Document decisions - Save to /memories/distributions/ for review
- Review regularly - Rebalance as conditions change
- Risk first, returns second - Preserve capital above all
Common Mistakes to Avoid
- Over-sizing on high confidence (still need risk limits)
- Ignoring correlation risk
- Taking on too many positions at once
- Forgetting to set stops
- Chasing after moves already made
- Panic selling on temporary drawdowns
- Revenge trading after losses
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