fpf-methodology

Guide AI coding decisions with auditable hypothesis-driven reasoning and Design Rationale Records.

1|Updated Dec 23, 2025

Search Tags:#audit #induction #design rationale #hypothesis #abduction #deduction

Ready to run

npx skills add https://github.com/ccf/claude-code-ccf-marketplace --skill fpf-methodology

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: fpf-methodology
Source: https://github.com/ccf/claude-code-ccf-marketplace/tree/main/plugins/structured-reasoning/skills/fpf-methodology
Command: npx skills add https://github.com/ccf/claude-code-ccf-marketplace --skill fpf-methodology

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill provides a structured, auditable reasoning framework (FPF) to guide AI-assisted coding decisions, ensuring transparent decision trails and reusable insights.

Core Features & Use Cases

Cycle-based inference: Abduction, Deduction, and Induction with Design Rationale Records (DRRs) to capture rationale and evidence.
Stateful knowledge management: Uses a .quint/ store for contextual context, hypotheses, verifications, and decisions.
Use Case: Teams building complex AI-enabled systems can rely on FPF to document decisions and justify choices during architecture reviews.

Quick Start

Use the FPF workflow to begin with /q0-init, then execute /q1-hypothesize, /q2-verify, /q3-validate, and /q5-decide to generate a Design Rationale Record.

Frequently Asked Questions about fpf-methodology

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

FAQPage Schema

How do I document architectural decisions and maintain an auditable evidence trail in AI coding workflows?▼

To document architectural decisions with an auditable evidence trail, use a hypothesis-driven reasoning cycle of abduction, deduction, and induction. This generates Design Rationale Records that capture context, verification, and final decisions.

What is the best way to structure hypothesis-driven AI reasoning for complex architectural reviews?▼

The best way to structure hypothesis-driven AI reasoning is running sequential workflow commands from initialization to final decision. This validates hypotheses through deduction and induction, producing a structured Design Rationale Record for team evaluation.

How does stateful knowledge management work for tracking AI hypotheses and design rationale?▼

Stateful knowledge management works by using a dedicated .quint store to track contextual knowledge states. It logs initialization, hypotheses, verifications, and final decisions, ensuring reusable insights and transparent decision trails across complex problems.

When do I need a Design Rationale Record for AI-assisted architectural decisions?▼

You need a Design Rationale Record when building complex AI-enabled systems requiring structured evaluation. It captures the abduction, deduction, and induction cycle, justifying architectural choices during team reviews with an evidence trail.

Can I use cycle-based inference for team discussions requiring structured evaluation of complex problems?▼

Yes, you can use cycle-based inference for team discussions requiring structured evaluation. It applies abduction, deduction, and induction to complex problems, generating auditable Design Rationale Records that document decisions and justify choices.