# Emergent Pattern Detector

> Detect when repeated signals or accumulated events begin to form a new pattern that should be treated as real enough to influence decisions. Use when separate observations may be converging into something more than isolated noise.

- Skill: `alubiama/emergent-pattern-detector` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add alubiama/emergent-pattern-detector`
- Raw SKILL.md: https://api.skillmd.com/api/skills/alubiama/emergent-pattern-detector/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: Alubiama (https://skillmd.com/u/alubiama)
- Updated: 2026-09-22
- Page: https://skillmd.com/skills/alubiama/emergent-pattern-detector

---


# Emergent Pattern Detector

## Goal

Notice when accumulated signals are turning into a real pattern.

This skill does not force pattern where none exists. It checks whether repeated observations now justify treating the pattern as decision-relevant.

## When To Use

- repeated signals are accumulating over time
- separate incidents may now form a meaningful pattern
- something feels like it is emerging but not yet settled

## Scope Boundaries

In scope:

- inspect repeated signals over time
- assess whether a pattern is emerging
- define what would confirm or break the pattern

Out of scope by default:

- final causal proof
- long statistical modeling

## Workflow

1. List the repeated signals or events.
2. Compare time, shape, and recurrence.
3. State the possible emergent pattern.
4. Judge whether it is decision-relevant yet.
5. Name confirming and disconfirming future signals.

## Output Contract

1. `signal_history`
2. `emergent_pattern_candidate`
3. `why_it_may_be_real`
4. `why_it_may_still_be_noise`
5. `decision_relevance`
6. `confirmation_signals`
7. `recommended_followup`
8. `next_step`

## Guardrails

- Do not promote coincidence into pattern without recurrence.
- Keep the threshold for "real enough" tied to decision impact.
- If the pattern is still weak, say so explicitly.