Forecasting

Apply quantitative and qualitative forecasting techniques. Use for demand planning, revenue projections, and trend analysis.

Security Scan

This is an instruction-only forecasting skill whose requirements and instructions match its stated purpose and request no extra credentials, installs, or system access.

Purpose & Capability

Name/description (forecasting, demand/revenue projections) align with the SKILL.md content: methods, selection guidance, and an output template. The skill does not request unrelated binaries, credentials, or config.

Instruction Scope

Runtime instructions are minimal and generic ('Apply appropriate forecasting techniques for $ARGUMENTS') and include a useful output template. They do not instruct the agent to read system files, environment variables, or send data externally, but the wording is open-ended and gives the agent broad discretion about what input/context to use—so you should be explicit when providing data and context.

Install Mechanism

No install spec and no code files (instruction-only). This is lowest-risk: nothing will be written to disk or downloaded by the skill itself.

Credentials

The skill declares no required environment variables, credentials, or config paths. There is no disproportionate access requested relative to forecasting functionality.

Persistence & Privilege

always:false (default) and no requests to modify agent/system settings. The skill can be invoked autonomously per platform defaults, which is normal for skills of this type.

Assessment

This skill appears internally consistent and low-risk because it is instruction-only and requests no credentials or installs. However, the instructions are intentionally general—when you use it, provide explicit, non-sensitive sample data and clear constraints (time horizon, required confidence intervals, allowed data sources). Avoid pasting secrets (API keys, raw database extracts), and consider testing outputs on synthetic data first. If you are concerned about autonomous runs, keep using user-invocation only or monitor the agent's requests for additional context before granting access to real datasets. If the skill later adds install steps, environment variables, or file-system access, reassess immediately.

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