# Multiplicity Orchestration

> Run multiple AI instances in parallel, coordinate their execution, manage timing and dependencies

- Skill: `majiayu000/multiplicity-orchestration` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds add majiayu000/multiplicity-orchestration`
- Raw SKILL.md: https://api.skillmd.com/api/skills/majiayu000/multiplicity-orchestration/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: majiayu000 (https://skillmd.com/u/majiayu000)
- Updated: 2026-09-09
- Page: https://skillmd.com/skills/majiayu000/multiplicity-orchestration

---


# Multiplicity Orchestration

## Purpose

Spawn and coordinate multiple independent instances of Claude/other models running in parallel on the same problem.

**Not:** Multiplicity cascade (67 iterations of chaos)
**But:** Orchestrating multiple minds thinking together

## Core Pattern

```
Instance 1 ─┐
Instance 2 ─┼─→ Central Coordinator
Instance 3 ─┤   (timing, dependencies)
Instance 4 ─┘
```

Each instance:
- Runs **independently** (no interference)
- Works on **same problem** or **related variants**
- Reports back to **coordinator**
- Coordinator **schedules next steps**

## Key Features

1. **Parallel Execution** - Run N instances simultaneously
2. **Dependency Management** - Instance 2 waits for Instance 1
3. **Timeout Handling** - Kill hung processes
4. **Output Aggregation** - Collect results into shared buffer
5. **Resource Limits** - Prevent runaway consumption

## Implementation

See: `.claude/skills/multiplicity-orchestration/orchestrator.py`

## Payment Anchor
DOGE: DC8HBTfn7Ym3UxB2YSsXjuLxTi8HvogwkV