fpf:propose-hypotheses

Execute a First Principles Framework cycle to generate, verify, and audit hypotheses.

Updated Apr 6, 2026
One-click install
npx skills add https://github.com/luicabref97/sushi-jungle-web --skill fpf-propose-hypotheses
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: fpf:propose-hypotheses
Source: https://github.com/luicabref97/sushi-jungle-web/tree/main/.agents/skills/fpf-propose-hypotheses
Command: npx skills add https://github.com/luicabref97/sushi-jungle-web --skill fpf-propose-hypotheses

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill automates a structured First Principles Framework (FPF) workflow to move from a raw problem statement to a documented, evidence-backed decision, reducing manual coordination and ensuring hypotheses are verified, validated, and audited before selection.

Core Features & Use Cases

  • End-to-end hypothesis lifecycle: scaffold a .fpf workspace, generate L0 hypotheses, verify logic to L1, validate evidence to L2, audit trust, and create a decision record.
  • Parallel agent orchestration: run verification, validation, and audit sub-agents in parallel to speed evaluation across many hypotheses.
  • Traceable artifacts: persist context, hypothesis files, audit reports, and design rationale records for review, reproducibility, and handoff.
  • Use Case: product teams investigating a sudden drop in retention can generate competing root-cause hypotheses, validate evidence streams, and produce a recommended action with rationale.

Quick Start

Propose hypotheses for why monthly active users dropped by 20% and produce an evidence-backed recommended decision.

Frequently Asked Questions about fpf:propose-hypotheses

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate and validate competing root-cause hypotheses for a product problem?

A first principles framework structures product investigations by moving from a raw problem statement to an evidence-backed recommended decision, generating L0 hypotheses, verifying logic to L1, validating evidence to L2, and auditing trust.

What's the best way to audit evidence backing for multiple decision hypotheses?

Auditing evidence for competing hypotheses is handled by parallel agent orchestration, running verification, validation, and audit sub-agents simultaneously to evaluate trust and produce traceable audit reports.

Do I need a repository workspace to run first principles hypothesis generation?

A repository workspace is required for first principles hypothesis generation to store .fpf artifacts, needing file read/write permissions to persist context, hypothesis files, audit reports, and decision rationale records.

Can I use parallel agent orchestration to speed up hypothesis verification and validation?

Parallel agent orchestration speeds up hypothesis evaluation by launching verification, validation, and audit sub-agents concurrently across multiple generated hypotheses.

How to create traceable decision records from raw problem statements?

Traceable decision records are created from raw problem statements by automating the end-to-end FPF lifecycle, persisting context, hypothesis files, and design rationale artifacts for review, reproducibility, and handoff.

When should I not use an automated first principles framework for product investigations?

An automated first principles framework should not be used for investigations lacking a clear problem statement or those without sufficient evidence streams to validate hypotheses and support a structured, auditable decision.