360 Secure SkillHub
360 安全技能中心

Forecasting

核心功能是根据预测需求匹配适配的定量/定性方法(涵盖移动平均、指数平滑等5类,覆盖短中长期场景),生成包含方法选择依据、历史数据对比、带置信区间的预测结果、假设及风险分析的结构化报告。典型场景有需求规划、营收预测、趋势分析,输入为预测诉求,输出为规范表格化文档,还提供验证更新等实用建议。

linuszz
linuszz
热度 51.61
最近更新 2026-05-11 17:34:32
分类:数据依赖:纯提示词应用场景:数据处理 / 报告撰写 / 财务

安装技能

复制下面的话术发给你的 Agent,它会通过 secure-skills 检索并安装该技能。

执行命令安装此 skill: npx secure-skills-cli add forecasting
Metadata
name
forecasting
description
Apply quantitative and qualitative forecasting techniques. Use for demand planning, revenue projections, and trend analysis.

Forecasting Techniques

Metadata

  • Name: forecasting
  • Description: Quantitative and qualitative forecasting methods
  • Triggers: forecast, projection, prediction, demand planning, trend analysis

Instructions

Apply appropriate forecasting techniques for $ARGUMENTS.

Framework

Forecasting Methods

Method Best For Time Horizon Accuracy
Moving Average Stable demand Short Medium
Exponential Smoothing Trending data Short-Medium High
Regression Causal relationships Medium-Long High
Scenario Planning Uncertain environment Long Medium
Delphi Method Expert consensus Long Medium

Output

## Forecast: [Subject]

### Method Selection

**Chosen Method**: [Method name]
**Rationale**: [Why this method]

### Historical Data

| Period | Actual | Forecast | Error |
|--------|--------|----------|-------|
| Q1 | 100 | - | - |
| Q2 | 110 | 105 | +5 |
| Q3 | 115 | 112 | +3 |
| Q4 | 120 | 118 | +2 |

### Forecast Results

| Period | Forecast | Lower Bound | Upper Bound | Confidence |
|--------|----------|-------------|-------------|------------|
| Q1 Next | 125 | 118 | 132 | 90% |
| Q2 Next | 130 | 120 | 140 | 85% |
| Q3 Next | 135 | 122 | 148 | 80% |

### Assumptions

1. [Assumption 1]
2. [Assumption 2]
3. [Assumption 3]

### Risks

| Risk | Probability | Impact | Mitigation |
|------|-------------|--------|------------|
| [Risk 1] | Medium | High | [Action] |
| [Risk 2] | Low | Medium | [Action] |

Tips

  • Use multiple methods for validation
  • Document assumptions clearly
  • Update forecasts with new data
  • Consider seasonality