What is the FPF Methodology for Claude Code?

The First Principles Framework (FPF) is a structured reasoning methodology that guides Claude through generating hypotheses, verifying logic, and gathering evidence to make sound architectural decisions.

When should I use the FPF methodology skill?

Use it for high-stakes architectural decisions, complex problems with multiple viable solutions, or whenever you need a durable record of why a specific path was chosen.

Does this skill make architectural decisions automatically?

No. Following the Transformer Mandate, Claude provides structured options and evidence, but the human developer always makes the final decision.

What is a Design Rationale Record (DRR)?

A DRR is a document produced by the /q5-decide command that captures the reasoning, evidence, and consequences of a technical decision for future reference.

How does FPF handle technical decision documentation?

It generates Design Rationale Records (DRRs) that include context, chosen hypotheses, evidence, trade-offs, and validity periods, stored directly in your repository.

First Principles Framework (FPF)

Name: First Principles Framework (FPF)
Author: ccf

byccf

0•

コラボレーションツール

Implements a structured abduction-deduction-induction cycle to generate auditable design rationale for complex architectural decisions.

The First Principles Framework (FPF) skill empowers Claude Code to act as a rigorous reasoning partner, moving beyond intuitive guesses to evidence-based decision making. By following a systematic cycle of hypothesis generation, logical verification, and empirical testing, it helps developers document the 'why' behind major architectural choices. This skill produces persistent Design Rationale Records (DRRs) and maintains a knowledge base in a local .quint directory, ensuring that technical decisions remain auditable, transparent, and resistant to cognitive bias as a project evolves.

主な機能

01Multi-phase reasoning cycle (Abduction, Deduction, Induction)

02Bias auditing and Weakest Link (WLNK) analysis for decision quality

030 GitHub stars

04Automated generation of persistent Design Rationale Records (DRRs)

05Local, git-tracked state management within the .quint directory

06Tiered evidence tracking from L0 Observations to L2 Verified claims

ユースケース

01Evaluating competing architectural approaches like database selection or API design

02Resolving complex, multi-variable bugs that require systematic hypothesis testing

03Creating an auditable trail of technical decisions for stakeholders and future maintainers

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add ccf/claude-code-ccf-marketplace fpf-methodology

For use in Claude.ai and ChatGPT

Download Skill