writing-ftd-documents

Drafts, reviews, and restructures Functional Technical Design documents using a ceiling-model template.

Updated Mar 28, 2026
One-click install
npx skills add https://github.com/thedutchvisiongroup/agent-skills --skill writing-ftd-documents-thedutchvisiongroup
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: writing-ftd-documents
Source: https://github.com/thedutchvisiongroup/agent-skills/tree/main/skills/writing-ftd-documents
Command: npx skills add https://github.com/thedutchvisiongroup/agent-skills --skill writing-ftd-documents-thedutchvisiongroup

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve? Writing Functional Technical Design (FTD) documents often produces either bloated, unreadable specifications or under-scoped drafts that miss privacy, security, and non-functional requirements. This Skill enforces a right-sized ceiling model: a small mandatory core plus only justified additions, with a mandatory clarification gate before any drafting begins. ## Core Features & Use Cases - Scenario-driven drafting: Classifies work as feature, project, or enterprise, applying matching size budgets, section toggles, and sign-off gates; enterprise designs ship as indexed multi-file bundles. - Review & restructure workflows: Audits existing FTDs against the standard (report-only), restructures legacy design documents into FTD format, or structures loose feature descriptions. - Compliance built in: Privacy-by-design and security-by-design statements are always mandatory, with DPIA, STRIDE+LINDDUN threat modeling, and measurable ISO 25010 NFRs for enterprise scenarios. - Use Case: A team building a patient triage assistant for hospitals asks for an FTD. The Skill identifies an enterprise scenario, runs the intake gate, and produces a multi-file bundle with NEN 7510 compliance evidence, DPIA decision, and measurable NFRs. ## Quick Start Ask the AI to write an FTD for your feature or project and answer its intake questions about scope, output language, and acceptance criteria format.

Frequently Asked Questions about writing-ftd-documents

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

FAQPage Schema
How do I write a Functional Technical Design document with AI?

Invoke the skill and answer its intake questions covering scope, functional needs, technical constraints, and five mandatory output questions (split, language, mode, AC format, filename). It then drafts a right-sized FTD with a mandatory core of scope, user stories, acceptance criteria, DoD, privacy, and security sections.

What is the difference between feature, project, and enterprise FTD scenarios?

Feature covers one story and one component (≤150 lines, no sign-off gate). Project covers multiple stories and teams (≤400 lines, Scope Summary sign-off required). Enterprise covers org-wide or regulated work and ships as an indexed multi-file bundle with DPIA, threat model, and compliance evidence.

Should I use bullets or EARS notation for acceptance criteria?

Bullets are the default for feature and most project scenarios. EARS (Easy Approach to Requirements Syntax) is recommended for enterprise, regulated, or safety-critical contexts where unambiguous 'shall' statements are required. The user always chooses in the intake, and one format is used consistently per document.

Can the skill review or restructure an existing design document?

Yes, Phase R offers three modes: Audit produces a prioritized gap report without modifying the source, Restructure rewrites legacy documents into FTD format preserving revision history, and Feature-structuring converts loose tickets or notes into FTD-feature format.

When should I not use an FTD document?

Do not use it for business cases, project plans, RFP/tender responses, quick single-endpoint specs, pure architecture diagrams, or user documentation. The skill explicitly declines these and points to the appropriate artifact instead.

Does the FTD skill support Dutch and HTML output?

Yes, output language can be English or Dutch, including Dutch EARS phrasing with 'zal'. Output modes are Markdown, self-contained HTML with sticky TOC and light/dark toggle, or both, with Markdown always drafted first.