draft-product-spec

Draft a stakeholder-ready features.yaml from product intent and optional vision.md inputs.

3|Updated Jan 25, 2026

Search Tags:#product specification #intent-driven development #roadmap scoping #features yaml #behavioral invariants #stakeholder artifacts

Ready to run

npx skills add https://github.com/kapilvirenahuja/garura --skill draft-product-spec

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: draft-product-spec
Source: https://github.com/kapilvirenahuja/garura/tree/main/core/components/skills/draft-product-spec
Command: npx skills add https://github.com/kapilvirenahuja/garura --skill draft-product-spec

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It turns product intent (and optional vision/roadmap inputs) into a structured features.yaml artifact that clearly defines what the product does and the behavioral rules it must honor—without locking in any implementation details.

Core Features & Use Cases

Create features.yaml from product intent: Produces a DRAFT features.yaml that captures product identity, invariants, scope boundaries, and vertical-slice features.
Enrich from optional vision and roadmap: Incorporates details from provided vision.md and roadmap artifacts for invariants and scope boundaries, while treating them as enrichment rather than gates.
Enforce “no technology” constraints: Strips technology names, SDK references, database products, hosting platforms, and deployment/implementation patterns from the generated output so it stays stakeholder-focused.

Quick Start

Ask the agent to generate a DRAFT features.yaml from your product intent using output_base .garura/product/roadmap/.

Frequently Asked Questions about draft-product-spec

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

FAQPage Schema

How do I draft a product specification from intent without including implementation details?▼

Drafting a product specification from intent involves synthesizing product identity, behavioral invariants, scope boundaries, and vertical-slice features into a structured features.yaml artifact. This approach strictly excludes technology names, implementation patterns, and deployment platforms to keep the output stakeholder-focused.

How do I create a features.yaml artifact for early product discovery?▼

Creating a features.yaml artifact for early product discovery requires synthesizing your product intent into end-to-end vertical-slice features. The process outputs a schema-conformant YAML file capturing the product scope boundaries and behavioral rules without locking in specific technologies.

Can I enrich my product specification using optional vision.md and roadmap epic artifacts?▼

You can enrich your product specification using optional vision.md and roadmap epic artifacts. These inputs provide additional context for defining behavioral invariants and scope boundaries, but they act as enrichment rather than mandatory gates for generating the features.yaml output.

What is the best way to define scope boundaries and behavioral invariants for stakeholder review?▼

Defining scope boundaries and behavioral invariants for stakeholder review is best achieved by structuring product intent into a schema-conformant features.yaml. This captures what the product does and the rules it must honor while stripping out specific technology references and deployment patterns.

Why does my generated product specification exclude technology and database references?▼

Your generated product specification excludes technology, SDK references, and database products to maintain a strict stakeholder focus. This constraint ensures the features.yaml artifact defines only the product identity and behavioral rules without prematurely locking in implementation details.

Does generating a features.yaml overwrite existing locked product artifacts?▼

Generating a features.yaml does not overwrite locked product artifacts. The drafting process specifically targets the output base path to write the schema-conformant YAML file while preserving any existing LOCKED artifacts within your repository.