# Identify Patterns From Repeated Observations

Teams often see repeated events but struggle to turn them into reliable rules. This skill structures observations, detects patterns, and reports confidence bounds with exceptions.

## Install

```bash
npx skillstore add bellabe/reasoning-inductive
```

## Metadata

- - Slug: bellabe-reasoning-inductive
- - Version: 1.0.0
- - Author: BellaBe
- - GitHub username: BellaBe
- - License: MIT
- - Repository: https://github.com/BellaBe/lean-os/tree/main/.claude/skills/reasoning-inductive
- - Ref: main
- - Supported tools: Claude, Codex, Claude Code
- - Risk level: safe
- - Quality score: 81
- - Quality tier: silver
- - Public page: https://skillstore.pages.dev/skills/bellabe-reasoning-inductive
- - Manifest: https://skillstore.pages.dev/api/skills/bellabe-reasoning-inductive/manifest

## Capabilities

- Collects multiple observations into a structured analysis frame.
- Separates frequency, correlation, sequence, cluster, trend, and threshold patterns.
- Forms candidate rules from repeated evidence and stated applicability limits.
- Adds confidence bounds based on sample size, strength, consistency, and recency.
- Identifies exceptions and conditions that may invalidate a rule.
- Suggests actions, monitoring needs, and further data collection.

## Use Cases

- Find Product Usage Patterns: Analyze repeated customer feedback, support tickets, or feature usage notes to identify reliable product behavior patterns.
- Review Sales Outcomes: Compare multiple deals to find recurring stalls, conversion signals, exception cases, and forecasting rules.
- Synthesize Research Observations: Turn interview notes, experiment logs, or field observations into evidence-backed themes with confidence levels.

## Prompt Templates

### Extract Basic Patterns

```
Use inductive reasoning on these observations. Identify repeated patterns, weak signals, and exceptions. Keep confidence levels conservative.
```

### Build Candidate Rules

```
Analyze these recurring events and form candidate rules. Include evidence, applicability, confidence bounds, and conditions that could invalidate each rule.
```

### Validate a Business Assumption

```
Use inductive reasoning to test this assumption against the supplied observations. Separate supporting evidence, contrary evidence, exceptions, and next validation steps.
```

### Compare Pattern Strength Across Segments

```
Analyze the observations by segment and time period. Identify stable patterns, drift, confounders, confidence bounds, and rules suitable for action.
```

## Limitations

- It does not prove causation from correlation alone.
- Small samples produce tentative guidance rather than strong rules.
- Results depend on the quality and consistency of supplied observations.
- It does not execute automated scans or access external data by itself.

## Best Practices

- Provide at least five comparable observations before asking for rules.
- Include negative cases and exceptions, not only successful examples.
- Ask for confidence bounds and invalidation conditions before acting.

## Anti Patterns

- Do not use this skill to explain a single isolated event.
- Do not treat correlation as proof of cause without further validation.
- Do not hide missing data or inconsistent observation formats.

## Security Audit

- - Safe to publish: true
- - Audited at: 2026-06-28T15:39:23.56\+00:00
- - Summary: All static findings were reviewed against SKILL.md and were false positives from Markdown syntax or reasoning terminology. The skill is a single Markdown guide with no executable scripts, shell commands, network access, filesystem operations, credential access, or prompt-injection language.

## Stats

- - Views: 152
- - Downloads: 6
- - Favorites: 0
- - Popularity score: 0