Most people can’t tell you what they believe in under 30 seconds. Not because they don’t have beliefs — they do, plenty of them — but because nobody’s ever asked the right questions in the right order. Worldview Synthesis is a Claude Code skill that walks you through a structured process: surface your beliefs, map the contradictions between them, and generate narratives you can actually use.
Install
/plugin install worldview-synthesis@2389-research
Then start a conversation with “Help me articulate my worldview.”
What it does
The skill runs six phases, each building on the last.
Bootstrap creates a project structure with data/ for structured belief nodes and narrative/ for generated prose.
Seed pulls ideas from your influences — books, people, experiences, anything that shaped how you think. These become the raw material.
Interrogate runs 4–6 rounds of multi-choice questions across domains most people avoid: death, money, conflict, ethics, legacy. The question bank covers 20 domains in 6 rounds, from mortality and spirituality through work philosophy and political orientation. Questions offer 2–4 genuinely different options, not leading ones.
Capture tensions is where it gets interesting. Every worldview has contradictions. You believe in collaboration and strategic ruthlessness. You value freedom and structure. The skill names these tensions and assigns a status: embraced (live in the paradox), resolved (found a synthesis), or unresolved (honestly don’t know yet).
Generate narratives produces outputs at ascending scales. A one-sentence core thesis. A mission that fits on a card. A synopsis. A full essay. Each one sounds like you, not like a template.
Iterate treats your worldview as a living document. Beliefs change. Confidence shifts. New tensions emerge. Come back and update.
Every belief is stored as a structured YAML node with a claim, confidence score (0–1), importance score (0–1), domain tag, and list of tensions it contradicts. This isn’t journaling — it’s a knowledge graph of what you think.
How it works
The skill uses Claude’s AskUserQuestion tool to run interrogation rounds interactively. Your answers get captured as structured YAML, analyzed for emergent tensions between responses, and fed into narrative generation. The whole thing stays local in your project directory.
Requirements
Claude Code with plugin support. No external APIs, no dependencies.